Wise and true words of Damocles, but for every vision there is a way. If you want to make this a success bad enough, it will happen. Just make sure that your vision does not deviate from the goal you appear to perceive. You can, of course not create an AAA title with this set-up, perhaps not even a B title or anything that has a future use other than gained experience for the participating members. And everyone must realise that, and understand why they do put effort in the game. And so that is the first task for you and the willing members to participate, to think of a way to gain more members, and how
to manage them.
I don't know what your plans are, but I represent thee my thoughts about said subject. They may be too far in depth for what you had in mind, but I think this is necessary to make this idea a success.
The easiest way to get members is to attract those who are new to the gamedesign business. They generally want to experience how it is to work in a team, and the project they join must offer this for them to continue working. These members do not yet care if their content will be of use in reality, yet will deliver quality that is below average, and may quit working quite fast.
Then there are the members who have some experience in creating work of quality. But often those members claim to have little time on hands. Well they do have time, but rightfully feel not like spending it too much on the project, as they gain little experience. Giving those a leading position is not always sufficient to keep them close to the team, and you, almost coincidently, need people with proper knowledge and enough spare time willing to participate. Those members are quite rare, but also important. It may be necessary to grab those outside of the community, as they do not particulary need knowledge about 3d gamestudio, rather about design, code architecture and art.
Filtering members on dedication, support, experience, newbie and uselessness will take time and may severely slow progress on the game. Members will cease to work and leave regulary, and the team loses more time in finding a new sufficient member. But eventually you should be able to manage the core of the team that will be able to make noticeable progress. From there, you should be able to operate the game course much more easily.
One last thing coming to mind is keeping the process alive. To keep things running it's wise to manage and document data, using SVN, a wiki, anything to keep information and core data available to all team members to read. Setting a short period course for clear milestones that need to be achieved within x weeks is vital, so everyone knows who works on what. It's okay when deadlines shift, as long as that is clear for everyone.
As a start I recommend you to create a main place for your future members to gather. Could be a site, a forum, an IRC chat, MSN even, but at least some private place for participants to speak. The next step may take some while, but that's generally not a bad thing. You can at least contact everyone when the basic structure is ready, meanwhile members can input their ideas.
Regards,
Joozey