Well what to say.. Fusion is a good idea. SS's release schedule is too vague to raise any excitment about it now. I realise that both programmer and player were excited to begin with, but without widespread dispersal of the examples, people lost a bunch of interest. The crazy thing is, even though it's going to be free, the lack of a marketing/advertising drive will be the game's ultimate dying point.
And like you said, MOD developers throw tons of energy into their mods and the mods become obsolete with the game. This is unfortunate, but it also seems to me, to be necessary because the new games add areas of growth in performance and gameplay. The new engines bring new game modes, interactivity with the environment, and even better multiplayer interaction. Note the massive number of modes in UT2k3 and now ut2k4 including a bunch of those mods that play WAY different than the stock game.
The biggest problem with WFF, is that it is being created by a single individual. He may accept input from others but for the most part, everything in the game will have some of his bias. Now this isn't a bad thing but it is commonly seen that a team can produce a much more balanced product because of the balancing act required between the team members. While SS was an active MODer and knows what Mod dev's need, it is also still possible for him to mistakenly limit things in the engine. I do realise that these can be fixed with it being opensource.
Finally. Just as there are distributions of linux, there are other engines around for people to use. Some, have visual superiority over the full quake3 engine. It would be nice for games to be made on these engines, but one big thing people, especially game programmers, do is lot things based on their primary use. An engine that is built and shown with a FPS game first is usually not considered as a first choice for being able to make a tetris game. For the engine to be the end all linux of game engines, it has to be able to make tons of different games; not just those that require a bsp tree and potential visible set.
Infernal