I don't disagree with this, but what you're missing is that CAP is more about the journey than the destination. If smogon was a game development studio, they'd probably have a team of a few people making pokemon concepts and then doing cycle after cycle of test/analyze/test/analyze. CAP is a creation-by-committee type of project, and that has certain disadvantages, but it also exists because it has a wide base of people who want to contribute to the creation, and subsequent testing, of a pokemon. The long, in-depth, sometimes heated debates are one of the benefits of the CAP process, so while they may be no substitute for empirical data, they are really the most valuable part of this process, and you'll learn things you never would have if you and a couple of guys were making / testing mons on your own server, for instance.
At some point we will have to make a decision between FPS and SPS, offense and support, or some compromise. It may not be the absolute best one, but it will probably be the best one as indicated by the reasoned arguments that have been made, and will at the very least be a reasoned one.