Disclaimer: This is a homelab question, but it reaches up into (IMO) the practical concerns of a SME sysadmin.

The drives in my homelab's bulk array aren't getting any younger, and they're starting to make that fact known. With the recent dramatic drops in the price/GB for larger drives, I'm looking to build a bigger storage array using 2/3/4TB as the atomic size, and about 20-30TB of usable space.

There seems to be a lot of concern over the statistical fact that as drive sizes increase, it becomes increasingly likely that any given drive will experience a URE when reading from start to end. Of course, that means that when you try to rebuild your degraded RAID 5, your MTTDL just became right-the-hell-now. Or at least as far as I've been led to believe. I mean, it's logical... right?

The discussion I've seen on URE probabilities for consumer drives seemed to focus on 0.5 and 1TB disks. Given the tone of those discussions, and the bigger drives of the now, is it a general consensus that RAID 5 rebuilds are destined to fail on big disks? Is RAID 6 really the solution, or are UREs on large platters becoming so common that running only two parity blocks is a risky proposition? What do you build your large-but-not-SAN pools on?

RAID is not backup. This is known, and not the focus of the discussion. All the irreplaceable data on my homelab is covered in a regularly tested 3-2-1 arrangement. I'm simply trying to think forward - this is my hobby, after all.

I'm looking for discussion and directions to reading material in addition to your advice. Please don't be shy!