Here is what I suggest should happen instead:
1. You currently have a list of suggested new games. I
image you'll be making a shortlist for the survey
fairly soon and the developers of these will be
contacted.
2. The developers of these games write to the docs
list, saying something like this:
"Hi. My game is being considered for GNOME. Please
could you take a look at the docs I have / help me
write a manual from scratch."
3. I strongly doubt that any of the games under
consideration would *completely* rewrite their UI
between now and freeze.
Most devs would have at least a vague idea of a
roadmap, and could say to us "features a b and c will
change, but the rest is stable".
EVEN with an app due for a complete UI change, we'd be
able to help them get a user manual template into
their CVS, and add to it an introduction and overview
of the rules.
With existing manuals, we can convert what's already
there to docbook, and help them get into shape for GDP
style. Looking at the manual for Hearts for example,
http://www.jejik.com/hearts/users/, it falls into the
trap of doing a listing all the menus instead of
focussing on tasks the user wants to accomplish.