As I understand it AdvFS and ZFS have somewhat different design philosophies. It's true that
both are intended for the enterprise market, but ZFS's philosophy seems to be that everything
works perfectly or you panic. In our experience ZFS fails hard on a number of typical failure
cases, requiring not just an fsck, but a complete rebuild before the disk can be mounted
again.
Most other filesystems fail somewhat more gradually... some disk blocks are lost, maybe an
inode allocation is lost and a file gets overwritten, or something unidentifiable is dumped
into lost+found, but overall the filesystem remains usable.