So with that thought in mind... would Open Source game development ever really work? Games are software, of course, but they are also a different beast. Games are created in a pipeline much more similar to films than other software development. Yes, they share similarities with software development (They are software after all), but in the entire scheme of things their entire development pipeline has always resembled other forms of entertainment rather than traditional software. With AGILE development, this pipeline has now become more of an amalgamation of the two ideas, but it's still a pipeline that very much requires an internal development team, and really cannot function by being spread out between many different people similar to how open source software works.
So what do you guys think? Could Open Source game development work with the various Agile and non agile pipelines in the industry today? If not, what would have to change about the traditional pipeline to make it work? How large would the core development team (if any) have to be in order to create something that others would be able to manipulate?
Additionally, I want to define that I am talking specifically about a game, not middleware or the underlying engine, or any other software used within the game. Needless to say, you can easily create open source software to use in the game, such as graphics engines, game engines, physics engines, etc etc, heck there are a few options out there now (OGRE, for example).
So could an open source game work? Why? Why not? What would need to change within the standard pipeline to allow them to work?