The Rubinius team should really be paying attention to the Google sponsored Python project Unladen Swallow and vice-versa. There is so much overlap in methodology they should really just be working together for half these things.
I'm sure each is fully aware of the other, or at least of the same source research into dynamic language optimization (which, I believe, comes largely from research into the Self language).
I (Evan Phoenix, author of the the article and project lead) talk with the Unladen Swallow guys almost daily. We've met up in person a number of times and have traded a lot of ideas, so I'm 5 steps ahead of you!
Unrelated question: is there a reason the docs in the git repo aren't rendered at http://rubini.us/documentation? They seem well written and it'd be nice to be able to browse it from the website.
The short version: with one thing and the other. I was working toward that at one point but it fell by the wayside. Good reminder, though, should get those up there. In the interim, of course, rdoc + doxygen should work for you locally.
Yeah that's what I'm currently doing. I'll be sure to submit doc patches if I notice any inconsistencies. Really enjoyed the blog post by the way. Thanks for that.
The MacRuby team is the one to talk to. This is Mac-specific Ruby on LLVM.