I pulled your email aside, as I'm on the road a lot, without access to
my notes. I have ran into a similar problem a while back and couldn't
remember until I got back to some of my notes.
This worked for me on a Debian system, and I'm guessing on most *nix
types. First I believe that zfs-fuse tracks its disks using a uuid
identifier, so if you change that on the drive, it thinks it is getting
a new drive. Second, and this may be over kill, I repartitioned my
"corrupted" drive and then put a new file system on it, then I
successfully reintroduced it into the pool.
This page gives details of how-to on Ubuntu, which works great with
Debian too. http://ubuntuforums.org/showthread.php?t=1378527