ZFS. GPL. Blah.
It's interesting that every so often questions like "how can we get more contributors to the kernel" seem to come up. It seems pretty straightforward. Any thread mentioning ZFS or GPL3 seems to attract dozens of never before heard of people. If only we could get them to review/write code.

Though the thing that really bothers me about a lot of the ZFS-on-Linux advocacy we're subjected to recently, is that there seems to be this misconception that the license is the only thing holding back acceptance into Linux, and that with a change of license, ZFS will zoom into Linus' tree with ease.

It took a lot of work to get another complex filesystem from another unix (XFS) into the kernel, and the results aren't particularly pretty in many parts. Even assuming someone crafted an abstraction layer for ZFS to work with the Linux VFS, it'd probably need to go through months of review before it got merged.

Given the shortage of filesystem code reviewers (see the reiser4 saga for example), the license is probably the least of ZFS's problems.