Isn't this all just evidence that "better" in most cases is a small margin? You can hate X, and prefer Y, but in most cases the X guys will finish their project. Methodology based crash and burns are pretty rare. And the things that are "not terrible" are not separated by that much.