The only way to get Durumn if you skipped her bundle is a month away, so I'm going to post about Feng first. Durumn's pretty damned snazzy though, so maybe you should grind that guild raid to try for her. Unless you're in a super pingu guild that wants everybody to spend gems on AP/revival, it probably won't stop you going for Feng anyway.

Not really a fan of the whole timeskip into OE thing Feng's got going on here - I like her 7* art a lot more and there's no way to keep it if you evolve her.

Feng's initial announcement and reddit discussion thread can be found here: https://www.reddit.com/r/bravefrontier/comments/5tz5r1/unit_details_blazing_phoenix_feng/

Anyways, I already did my pulls for Feng and got lucky, but if you're waiting for my input before rolling for her, here it is.

Stats/Arena/Animations

• Feng's pretty much an all-rounder like her parents, with no particularly special stats at the outset. That said, she has slightly above average HP and gets a 50% enhancement option for it which will make her pretty bulky if you choose to take it.
• As for the arena, Feng's a case where she has a bunch of important arena abilities, but no damage pumps (beyond generic-tier 60% ATK) or BC gen to back them up with. 15-20% mitigation and a chance-based angel idol on her LS, threshold angel idol on her ES makes her seem like a really good arena leader, but if we compare her to the big three global leaders, she offers less damage than Terry or Galtier and no BC gen to match up to Juno-Seto - her only advantage is the mitigation which doesn't apply to some of the hardest hitting units in the meta. She looks better numbers-wise than Eleanor, but Eleanor's only really used for the bonus ABP/CBP.
  • That's not to say she isn't useable, but if you picked up any of global's LE arena gods you don't particularly need her.
  • Speaking of Eleanor, trying to use her cost me a 2500-ish win streak in regular arena as her lack of damage failed to kill any enemy units turn one and a Kaz-rn'd Mifune just one-shotted everybody. Wonderful. At least there aren't any achievements beyond 500.
• Feng's animation is blanket-ish, but not perfect. That said, all of her delays are 3 or 6 frame gaps, meaning she can be covered by a single blanket. The entire animation covers 102 frames, which is about 53 less than a 45 hit spark blanket, giving plenty of wiggle room. Feng's also a move speed 3, type 3 unit meaning she can be dupe sparked easily.
  • Her normal attack and BB's hits include every single hit frame used by her SBB, meaning if she fails to fill for some reason or you swap her mode to BB only for a turn for the buffs, your spark blanket should still cover all of her hits. Neat.

Passive Skills

LS

• Feng's LS is seemingly the biggest subject of discussion in her announcement thread, perhaps for good reason. It offers 60% baseline HP, 60% ATK/DEF/REC based on HP remaining, 15% mitigation against Fire, Water, Earth and Thunder units and a 20% chance to not die when you are killed (limited to one death prevention per fight).
  • The first thing that comes to note is the stats based on remaining HP. This makes Feng the rather counter-intuitive case of a defensive LS that gets worse the more damage you take. I personal think with the limited scope of abilities on this LS the stats may as well have just been standard with no scaling factor. Some other important notes include:
    • Passive stat scaling cannot go beyond its limits in the way HP-scaled BB skills can. The calculation only considers current HP and actual max HP, so it caps out at 60% ATK/DEF/REC.
    • This bonus does stack if you use more than one of it, despite having a buff icon.
    • When a unit's life is saved by the chance-based angel idol effect, that unit stays at 1 HP (unless otherwise healed). This means units saved by chance-based angel idols will not even have the REC bonus to help them heal up.
  • The mitigation here is particularly strong at 15%, but the main cause of discussion is the very expensive enhancement that ups it to 20%. At 20% mitigation you can stack two Feng LS with BB mitigation and 10% mitigation from spheres or buff-trigger ES/SP to make units effectively invincible, with some caveats like the enemy being a correct element, not buff-wiping you and not ignoring mitigation. My personal take on this:
    • 5% extra mitigation sounds small, but isn't - with the minimum expected BB, it takes you from 65% mitigation to 70% mitigation, which actually works out to a 14.3% reduction of damage taken. With two Fengs, both taking it, you go from 80% mitigation to 90% mitigation, which is a 50% reduction of damage taken. In other words, the more mitigation you have, the more powerful stacking even small amounts of additive mitigation becomes.
    • On the other hand, it's very expensive, and basically gimps your Feng in any other role as she cant afford all of her nuking kit or utility kit alongside it.
    • A few are complaining that it doesn't cover Light and Dark. You know what elements are going to be present unless you're doing blind runs or guild raid, so it's a non-issue in my mind. If she won't work wait for OE Cardes.
  • Some important mechanics for mitigation, as a refresher:
    • Passive elemental mitigation stacks additively with BB or SBB tier non-elemental mitigation buffs, buffs spawned from passive effects (first x turns, mitigation after taking x damage, mitigation after collecting x BC, and so on), and all elemental/non-elemental mitigation passive effects on LS, ES, Spheres and SP enhancements.
    • Passive mitigation has a hard cap of 50% which is only bypassed by adding mitigation buffs.
      • With the exception being chance-based 100% mitigation, not relevant to Feng.
    • Elemental mitigation only looks at the attacking unit's base element - the actual element of the attack itself is ignored.
  • Finally, she has a chance-based angel idol. This gives us a 20% chance that our units will not die when they are killed. This part of Feng's LS can only trigger once within a given fight or wave, making it considerably less useful in long content than in the colloseum.
    • Once triggered this effect lasts the entire turn and can probably only be bypassed by attacks which directly kill the unit (such as Kalon's stuff in his EX dungeon). These are rare and the case untested, though.
    • As mentioned earlier, units saved by this will be at 1 HP, which will matter for the sake of HP-scaled effects. It's possible the units will be partially healed by Heal-when-attacked effects, however.
    • This effect does stack, but stacks by having very instance of it attempt to proc whenever a unit dies. This leads to cases where a charge of chance-based angel idols can be wasted by having multiple rolls succeed, but at least the chance of surviving the first death increases as a result.
    • There's no way to keep track of how many chance-based angel idols have been consumed in any case where a unit has more than one angel idol source (in cases where a unit gets multiple angel idols from a single effect as commonly appears on ES effects, they can't multi-trigger as the efefct only rolls once). Once you've seen a unit survive once, avoid making assumptions that it can take another death.

ES

• A simple two-parter. Feng's ES gives her some heal-when-attacked and a threshold-based angel idol.
  • Heal when attacked is a somewhat unreliable form of mitigation which, at a low chance restores a low amount of the damage taken. In this case it's worth roughly 5.625% mitigation against final damage taken.
    • Note that this type of healing doesn't kick in until the animation causing the damage completely finishes, and only if the unit taking the damage is still alive at that point.
    • In Feng's case she may still be alive after taking lethal damage due to her leader skill or ES angel idol. In the former case it's possible for her to heal from the damage she's taken to get that far and it'll actually mean something going into the next turn as her SBB is HP-scaled and her LS has HP-scaled stats. In the latter case it doesn't matter; threshold angel idols fully heal the unit after they trigger.
  • Threshold-based angel idols are effects which apply an angel idol buff to the unit when it passes a certain HP limit. This one, as is the norm, is triggered when Feng passes 20% HP. Some notes:
    • It's possible for this kind of angel idol to be triggered without consuming the buff if the unit goes below the threshold without actually hitting 1 HP. In such a scenario it can be buff-wiped on subsequent turns.
    • The angel idol will be skipped if Feng goes from above 20% HP to 0 HP in a single hit.
    • If the angel idol is skipped and Feng is saved by a chance-based angel idol such as the one on her LS, her ES angel idol will be triggered too (because she's below 20% HP). Similarly her ES angel idol can be triggered first and when she hits 1 HP her LS angel idol will attempt to proc anyway - unfortunately, you have no way of knowing if that was the case.

BB Skills

BB

• A regular AoE which gives a permanent HP buff, 3 turns of HP->DEF conversion and heal-when-attacked, and has a 10% chance of reviving dead units.
  • At 25% it's tied for the biggest non-UBB HP buff we've seen at baaseline, but sadly lacks an SP enhancement to match Ilm's fully enhanced 30% HP buff. As an HP buff it has a degree of permanence, only disappearing if you use a larger HP buff or enter a map screen in GQ or Raids, and potentially even carries over to units in the same slot in a trial.
  • The HP->DEF conversion ratio is 35%, giving 3500 DEF per 10000 HP or around 10k DEF for a decently equipped end-game unit (around 200~230% HP). By comparison equivalent ATK->DEF converts need you to reach around 12500 ATK, which is around 200~300% ATK depending on the unit, to match that. This is still relatively trivial if you bring ATK buffs and all-stats spheres/LS/elgifs, so Feng's convert is not the best available.
    • That said, there are LS pairs which might not boost ATK at all, or you might not have the flexibility to slot an ATK buff, and at the very least you can expect a baseline HP boost of around 180% in any squad suited for endgame now, making HP converts easier to work with. It's also not susceptible to ATK down and Injury.
    • Note that convert buffs cannot be re-converted to another stat to prevent infinite loops, though there wouldn't be any possible at the moment as there's no stat->HP conversion.
  • The heal-when-attacked is just the buff form of the effect I already taked about on her ES. As I didn't mention it and it's relevant to her, though, just a quick note on how the effect stacks - the game doesn't add the numbers together, but instead rolls each heal when attacked chance individually, so a unit can heal from both, one or neither if it has two sources of the effect. This means there is value to taking multiple sources that stack, but you can't hit 100% chance or heal rate using effects that don't do that inherently.
  • Finally, the revive. This is a flat 10% chance that dead units will be brought back to life. Feng also heals the unit to 20% HP, which probably isn't enough to survive, so you'll need to patch it up. Revived units don't receive buffs for about 1 second after the revive is triggered, so unless you're firing Feng early to check for revives your revived units aren't going to have mitigation or DEF buffs (so be careful with it)

SBB

• An HP-scaled AoE which gives 3 turns of HP->ATK and 450% BB ATK, burst heals for a decent amount of HP and also has a chance to revive dead units.
  • The SBB is HP-scaled, meaning the higher Feng's total HP% is, the more damage it can deal. This kind of effect has a damage boost scaled as (Current HP / Max HP) * HP%, so as an example at 100% HP and 16000 current HP out of 20000 the damage boost will be multiplied by (16000 / 20000) * 2.00 = 1.60. This allows HP-scaled effects to easily reach the base damage cap of 99999, with Feng being able to hit it at as low HP values as 150% while at max HP (two leader skills and a sphere, basically). For future note, Feng hits the damage cap at around 2350% ATK as Lord.
    • Some people mentioned in teh comments they wanted to know what her damage modifier here is; it's the standard 200% base 700% HP scaling seen on other units (frequently written as 200-900%).
  • The HP->ATK is at the same value as the HP->DEF on her BB. This means a properly supported DEF->ATK buff will, again, be stronger, though most unit's DEF stat is low enough that it's at least a closer match than with the HP->DEF before. The ATK granted by this buff is used for normal attacks as well, making the buff considerably more valuable if you're doing any hitcount stuff (clarification: that last part applies to any stat->ATK conversion)
  • BB ATK is a relatively common buff but the top tier value of 450%, with no SP enhancements required, is pretty rare and valuable, being one of the most powerful damage boosts you'll be using in the kind of content where Feng's a desirable lead or sub unit (except arena). That said, the upcoming competition for this buff is mainly found in a particularly OP unit and the clash with Regil here may actually hurt Feng in the long run.
  • Feng's burst heal is particularly large at 3500-4000 40% of Feng's REC 100% of recipient's REC, which is decent for an attacking healer. But it has one more hidden feature which makes Feng a real unexpected gem for damage-dealing - the heal applies on frame 0 of her attack animation, while her damage is calculated on frame 1. As her SBB is HP-scaled, this means Feng will heal herself to full before calculating damage, and almost always deal damage based on her max potential. On the other hand, the heal will always apply before damage reflect unless you intentionally delay Feng's action so only she gets hurt by it.
    • And yes, I took some time to test this, taking a Feng with effectively 1400% BB ATK at 1% HP in the battle simulator and having her SBB deal capped damage due to healing 20k HP as it fired (increasing her BB modifier by around 1400% to the cap value). Fengs with less buff support may not have as strong a heal or base damage output to always allow this, but still - her heal triggering so early does a lot for her damage output in ad situations and makes her a lot more stable than other HP-scaled units.
  • The revive here is the same as her BB. Unike Juno-Seto, Feng attempts to revive on both BB and SBB, ensuring it's easy to fire every single turn if you really need to. It has a slightly lower proc rate, but is much easier to use consistently.

UBB

• Feng's UBB is a non-attacking type which gives huge defensive buffs such as: 100% mitigation for 2 turns, 35% HP, 60% HP->ATK for 3 turns, full HP regen for 3 turns and a 100% chance of reviving units at 20% HP.
  • 100% mitigation is a very strong UBB factor which prevents units from dying to anything which mitigation applies to while active. There's not much else to say here, but most squads want UBB mitigation available and having a non-mitigator option for granting it is particularly helpful in BC-starved environments.
  • The HP buff is just a bigger version of the one she has on BB. HP buffs don't follow the normal rules for buffs - UBB buffs don't stack with BB buffs, and instead when multiple HP buffs are used the largest one is the one which stays in play (both of which are different to normal buff behaviour, where normally BB/UBB stack and the newest buff stays). Again her buff is smaller than the strongest UBB HP buffs seen, but you're not going to be using he UBB just for it and in fact are likely to not be using those UBB at the same time.
• The HP->ATK is just a bigger version of her SBB buff again, though this time it stacks. While there are some DEF->ATK convert UBBs which blow this one out of the water, the same thing as the HP buff applies - you're unlikely to be using them at the same time.
• Full HP regen is just that. HoT effects counteract DoT by applying after it, so this effectively ensures that your squad won't be taking damage from any DoT effects that normally bypass mitigation.
  • Though there is a scenario where a mob could have damage reflect and DoTs active and your squad could die to the DoTs before the regen finishes. I guess unless it's global exclusive content that does that we'd have a lot of forewarning, though.
• Finally, a 100% chance revival - the first of its kind, actually. Unlike all other revive UBB, Feng will never fail to pick up your squad, but this does come with a drawback - units Feng revives will only be at 20% HP (other revival UBB typically revive at 100% HP), and none of Feng's other UBB buffs will be applied to the unit, nor any buffs used within ~1 second of her UBB.
  
  • In other words, after using Feng's UBB, wait for the animation to finish before using any other units, and make sure you have heals and defensive buffs ready to top off the revived units. Luckily her UBB is non-attacking, so waiting for it will not be a damage loss.

SP Enhancements

Good ol' controversial SP pricing. Well, not so good. She could really do with a 5-10 point adjustment on some of these. I also feel like she's missing a damage cap increase which would actually matter for her where it didn't for many recent unit releases.

• 20 SP: 50% HP
  • Feng starts off with a really strong enhancement for her survivability and base damage. 50% HP works out to 350% SBB mod for her, as well as around 4750 to 5500 HP depending on her type. Most builds will probablyw ant to find a way to include this.
• 10 SP: 80% boost to Atk relative to remaining HP
  • On a unit... which reaches the damage cap on SBB really easily... there is a relatively minor ATK boost.
  • If you're using her in the arena, though? Good stuff. 80% ATK on turn 1 is nice if she doesn't take damage before then.
• 10 SP: 80% boost to Def relative to HP lost
  • Or you could take the inverse, a DEF boost that improves the more damage she takes. The effect over each turn wil be relatively minor, however, and overall I don't see this being too reliable. I'd rather have the ATK boost personally, even if it only affects BB and normal attacks.
• 20 SP: 80% boost to Spark damage
  • Not quite as cost effective as some competition provides, but still a solid damage boost.
  • Unfortunately using her LS build fores you to choose between this and her HP boost, which most people would prefer. That said, she can cap damage relatively easily without the HP boost; it's more a matter of survival than anything.
• 65 SP: Enhances LS's reduction to Fire, Water, Earth, Thunder elemental damage [ 5%] effect
  • Feng's most controversial SP enhancement. I discussed this way back in her LS section, so there's not much more to be said about how effective it is.
  • That said, I'll say that 65 SP is basically 5 too many for an HP/Spark Damage build, which is annoying enough, but also 10 too many to take one of her 25 SP buffs with the HP. Taking this thing arguably ruins the rest of her build for a better LS.
  • And the cost isn't very competitive at 65 SP comapred to Regil's 30, either. You may still want it for what it is, but gumiplz.
• 30 SP: Add 1 turn Stat Debuff Immunity to BB/SBB
  • Adds a relatively uncommon immunity buff to Feng's BB and SBB. That said, as you're unlikely to be using ATK->DEF or REC->DEF conversion with her, stat debuffs are less threatening than they would be to another squad (not that they're particularly threatening to begin with)
  • Basically a utility option. Probably not worth losing the HP passive for if you're set on taking her LS enhancement, however.
• 30 SP: Add 1 turn Crit Immunity to BB/SBB
  • Another immunity buff, this one being signficantly more important in the niche occasions where it can be used.
  • Most squads want something like this. Feng doesnt have too much clash with certain other units like Ark which frequently take this enhancement, but it's an option regardless.
• 25 SP: Add considerable Atk boost [120%] for Fire, Thunder types effect to BB/SBB
  • Adds a Fire/Thunder only ATK boost to her BB and SBB. This ATK boost will stack with any non-elemental ATK boost, and as such is good for strategies involving ATK conversion.
  • That said, Feng provides a stat->DEF conversion of her own already. This makes ATK boosts relatively unimportant to squads using her other than a source of damage (unless you're using an ATK->REC conversion for Rahotep or something strange like that).
  • Note that there's no included DEF boost, unlike several other units with elemental ATK buffs in their kit.
  • The cost is rather annoying at 25 SP, not allowing Feng to take it with her LS enhancement while also taking the HP boost.
• 25 SP: Add considerable Atk boost [120%] for Water, Earth types effect to BB/SBB
  • The water/earth mirror of the above option. The same things apply.

SP Builds

In general: LS build, Max Damage build, Utility build.

For the LS build, the most common choices are going to be 50% HP and ATK based on HP remaining or DEF based on HP lost. That said, you can instead opt to take a single buff from the 4 available, and if that buff is one of the ATK buffs you can take the ATK or DEF passive too.

The damaeg build takes all of the ATK buffs to make sure Feng works as well as possible in any squad. Her LS is still usable without the enhancement like this, but the intention of the build likely means you're taking her as a sub unit in a squad with a damage boosting leader.

For the utility build, while I feel the crit immunity should be taken by default, you'll have a choice of either taking the debuff immunity with 20 SP of passives (recommending 80% spark with this option), one ATK buff with the same choice of passives, or both ATK buffs. If you go this route it's really up to you, as Feng's nature as an HP->DEF convert unit make all of these buffs somewhat less significant than they could otherwise be.

Conclusion

Feng's... interesting. Her SP options are, frankly, annoying and many players have already voiced concerns over them at this point. She loses so much going for her LS enhancement, but her LS enhancement can potentially make it trivial to break the game... but it basically denies you other fun stuff she has available. So, annoying. Mostly caused by pricing.

Her leader skill, if enhanced, offers one of the best available survival options for endgame content. However, it does very little else, and I find that to be an important weakness when her competition includes units like Regil for quest content and units like Juno-Seto/Terry/Saltier for the colloseum.

Then there's her skill kit - between BB and SBB she has two somewhat subpar converts, decent HP and BB ATK buffs, a low value heal-when-attacked buff but an absurd combination of HP-scaling with a heal that applies in time for damage calculations, and a revive that - unlike any other unit with the effect - appears on both BB and SBB. And her UBB is ridiculous with a 100% revival rate and 100% mitigation/regen, only particularly being held back by the nature of revives not allowing buffs to apply from the same skill.

But for everything her BB/SBB kit has, the clash with better converts puts her in a weird spot for squad-building, as she can't work well with several popular units because her converts and HP make up the most desirable factor of her kit (just to contrast, Rain has 80% ATK->DEF and up to 20% HP, on top of other very desirable buffs such as tristat and mitigation). The other stuff she can bring is nice when it matters, but everything in a way is niche or common here.

So making proper use of Feng is going to present some difficult choices. I think she can be good if used well, but you have to tailor your squad to Feng specifically. At least it's not as bad as Rahotep in that regard.