I agree wholeheartedly, but I still feel like the lack of available
documentation is greater than the lack of quality documentation tools.
I think this would be an *awesome* summer of code project or something
like that, so long as serious efforts were made to plan it properly,
but for something like Ruby Mendicant, I think I'd be more useful for
my writing skills and ability to dive into undocumented code bases
than I would for building a documentation generation system.
I guess my point is that in my day to day work, I've felt the itch of
needing better documentation tools, but never really had an urge to
scratch it.   That's why I don't think we'll see this as a candidate
for a project I can work on, but I definitely would back an effort to
do this, perhaps in Summer of Code or another way.  It's a big project
that deserves funding, for sure.
-greg