Just found this thread - I've got a few thoughts generally speaking, feel free to take them or leave them:
Definitely go open source. Not just because of the warm fuzzy feeling you get from releasing the source code to the world, but because it makes it easier for people to join in, it means you can get as much free storage as you need and perhaps most crucially you get to use GPL licensed libraries, of which there are *loads*. Of course, there are commercial equivalents, but then you're just shelling out money for equivalent libraries with less support and for something that (let's face it) is very unlikely to turn into a commercially viable project.
Looking back through those in this thread, there's a lot of people involved who are just starting out with the development side of things, and perhaps aren't all that confident. So in order for this to work well it should be easy to get going, easy to use. In that regard I'd plump for using Google code over sourceforge or github (the interface is much easier, sourceforge is a beast and github is, well, bloody github.) Stick with subversion, I know all the rage is with DVCS these days but we're not planning on doing loads of branching at present and can always switch later to Hg or Git if we so wish. Svn is stable, well used, has loads of resources around and more well developed clients (like tortoise) than you can shake a stick at.
When you get a Google code page (or whatever else) then use it. Use the wiki pages for discussions, use other wiki pages for finalising ideas, create FAQs and instructions for people getting the code and modifying it, requesting commit access and so on. Make them easy to follow.
In terms of hardware - if you want this to be as accessible to as many people as possible, then cut any specialised hardware out of the equation, at least for now - and by that I'm including things like Arduinos that are readily available. Not everyone has them, not everyone even wants to get them and it adds a whole other layer of complexity that can put people off. As well as making things more difficult of course.
It may seem like I'm being a bit of a kill joy and condemning all "interesting" and "cool" and "new" things here, but as said to start with, the majority of these things don't last - the interest dwindles substantially just after the work-load becomes apparent, people become busy when they don't expect it, and then the thing falls apart. I'm not just talking about projects here, it's the same everywhere. And this is especially the case when you're looking to create something for the sake of learning and creating something, rather than because there's a particular need for that product. So in order to make this succeed it does need to be made easy to get going, easy to carry on and without unnecessary hurdles.
Also, make sure everyone is happy with what's being chosen, don't plough on ahead too soon. Just in this thread so far we've got some people going as far as discussing the particular features they want in a blog project, and others that haven't decided on a blog at all. Those are already early warning signs of a fragmented project where people's attention starts to wander quickly... take your time, there's nothing to be gained from rushing into this.
Now regarding my participation - I'll happily do my best to answer any questions that come up, but I'm afraid I'm way too busy at the moment to commit to something like this on a grand scale. I will however, if you wish, set up a Google code page and point people in the right direction with those sorts of things. But before I (or anyone else) does that, decide unanimously on what you're doing, decide on a project name, and make sure everyone is happy