There wasn't a destroyed pool, it's the harder version of trying to rewind time on the filesystem. It's worth trying once the disks are fully backed up, but it's fussy enough that I can understand why they made it plan B.

It didn't make the cut for the article, but I do recall trying to import older uberblocks still in the label ring buffer with zpool import -T at some point, but none were before the zfs destroy. I assume that was because TrueNAS logs to the .system dataset and presumably wrote through the entire ring buffer before I got to it.