what happens is the coders get ahold of things, the project is in there hands, everyone waiting around to do something, and in the end nothing happens. this approach has been tried twice with no good outcomes.
i still firmly believe that it should be a project of several smaller games... which prolong the story of the characters, fleshes everything out, have many basic parts down like theme music, character likenesses, and a back ground established, then proceed to make the bigger game.
the story should be shaped by the majority of people in the project. so that the story direction and theme most people agree on. once the confines are all laid out and agreed upon people work within them confines. for example lets say there is a sampson like character, and someone makes a mario type game side project. would prove to be very intresting and funny. then maybe someone takes the opposing side and makes a tower defense game, to stop the invaders. some one makes an rpg, another does a strategy.
one thing we dont have is lots of money or media catching ability. however if we start a general story and have many different game types it wont be a stand alone game but part of a group of games, storys, songs and other things. your characters instead of being random and new, have meaning, possibly a very deep story behind them, a better deeper explenation of the surroundings and motives of characters. and the ability to take them characters and put them in different settings and places. you may play with a character when there 20, and end up playing with them again when there 70. starting to play them as a solidier then play with them as a king or general. switching from fps, to rts.
the problem is making to narrow of a story, or killing all the bad guys to quickly, and leaving none behind. or writing the world as to confined restricting other game possobilities.