I was thinking along the lines of a Death Match Alien Breed ?? I mean that project is already on going, and I dont think it would be to hard to add a death match server to it.
The lack of multiplayer games in general is a real shame as well. Or maybe theres a larger problem here. IMHO the problem is, when developing a network game, the network has to be at the core in the beginning. Its difficult to add it in afterthought. Many Blitzers start dev'ing without network in mind and Alien Breed started without one as well.
GNet is merely a server registering service and I developed many versions similar. There may be an issue with GNet's Refresh Delay. Its not a problem for a turn-based game or real-time application were fast update is not critical, but, it can be trouble for twitch games. Theres always a workaround.
Community Projects are tough. I found the problem to be in the community itself. Its chock full of developers, who are are striving to create THEIR game, not yours. I have struggled for years by myself with various Game Development Tools, Authorization Systems, and Languages.
After many years I've realized 1) that most games have common features. 2) If I concentrated on one feature I could perfect it. I'm not convinced that OO is the only way that large cool games can be created. But hell thats why I like BASIC.