I've been looking at it a bit closer, and the main problem with the site (apart from parts being broken and a difficult CMS, according to z98) is the fact that there is much redundancy. It needs a big cleanup. The info could be merged with the FAQ, the links that link to pages that show almost nothing be deleted, Icons with links for the development be grouped together and being cut down when it's redundant or obsolete, etc.
All that pertains to the development should be put there, and not mixed on other pages. Broadly interested and curious parties should find all they need to know on the first page, with some nice titbits of (updated) news. Then an area for the bit more in-depth interested persons who want to try out, and one specifically for the devs.
It could use an overhaul too, but it's not all THAT important, compared to the rest. Some new graphics with shadow and rounded corners would be nice, but I thought there were already some nice examples when the site-overhaul was first discussed in 2009 (which apparently went nowhere). The main theme should be kept throughout the site, though: white and variances of blue/grey. The site must have a uniform feel to it.
This is the most arbitrarily chosen decision. Is it still the prevailing wish to go for Drupal? I'm not sure if it's all that a wise a decision, but I guess one has to start somewhere. As it is now, even an ordinary site with html5 and CSS would already be better maintaining-wise, if I understood correctly. A CMS is all fine, but it's better to have none than a bad one. Of course, a new forum (and underlying mechanism) would be highly recommended too. All in all, you have a lot of possibilities to choose out, these days.
Well, so far my suggestions. The main reason it's gone nowhere during these last years (from 2009 to now) is mainly because different people have to agree on different things during a considerable time. Neither the first nor the last requirement is easily met, where ego's and different views may collide and other r/l demands take peoples' time. And if it's not an open project, things stagnate once enough people 'disappear'. It would therefore be better, imho, that such a project/overhaul would be tried in incremental steps (publicly accessible and viewable), as a true open source-like project. That way, it never really 'vaporises' into thin air, and as long as there's a small dedicated amount of people working on it, it gets better, up until the point it's good enough to transit to completely.