Well... It's because of all of that; that WFF looks like a great arcation point for building an actual engine project, not just a game. By 'Linux Of FPS Engines' I don't so much mean something that will take numbers and kick ass on features and capability as something that has one Programmer in charge (guess who?) that would steer, manage and moderate method and direction.
Linux was never intended to be a State Of The Art OS, but as an Open OS built around a legacy and uniform design that anyone could contribute to; that it became a SOTA OS is beside the point, which is that it attracted a lot of talent and became the locus more for creating an 'Encyclopedia Of Recipes' (to pursue your analogy) then get all the cooks in one kitchen and baking one pie.
Perhaps UNIX and it's inherently modular design does not make a fair comparison to 3D Render Engine design and development and all this is wishful thinking. It just appears to me that there there is a lot of duplication and waste of effort on BSP engine development, and that a lot of game and mod developers would appreciate a 'single qualified and standardized source project' that they could learn build on and extend, rather then playing 'wait & see & relearn' with next year's crop of commercial engines.
Perpahs I should bark up other trees.
:O)
.