When and why to deprecate filesystems

I think it's highly unlikely that anyone will add COW to the EXT filesystems.
These days, the EXT filesystem is the 'ol reliable' of linux filesystems. It's stable, easy to recover from crashes, simple enough in structure, and an excellent compromise between keeping data local and spreading things out. It's very low overhead: at the end of the day, you just find the block and change it. COW, even just of individual files rather than all metadata, requires a lot of additional work for every operation. For instance, deleting a file doesn't just mean removing the extents, but it also means checking other files to see if they're using those extents rather than just marking them as free. You also lose the nice, low-fragmentation behavior on hard disks where seeking is costly: file blocks stop tending to be in nice ordered segments, and start tending to be chaotically placed across the disk. These aren't properties you want in a filesystem that is trying its very best to be as stable, reliable, lightweight, and boring as it can be.

Ext4 is so reliable that even if the first 600 MB of your workhorse laptop's disk are overwritten with, say, the Debian Testing installation ISO because you wrote of=/dev/sda instead of of=/dev/sdb, you can recover almost all your data and have a fully functional system again with a bit of effort, some abuse of fsck, and a bit of quality time with debsums. I don't think I can say the same about BTRFS, though I can't say I've tested that theory.