Yeah...
This blog post by Bryan Chung I think is absolutely relavant here. It is really the same kind of thing. Elite gymnasts are buff so the way to get buff is bodyweight training or "crossfit" style training. Or elite sprinters are muscular compared to skinny long distance runners so....
While there is a kernel of truth that certain ways of training are good as compared to others what we don't realize is that the sample size is SO SO small....He calls it the "sport causality bias", or "elite athlete selection bias".
Well, he explains it better than I could:
http://evidencebasedfitness.blogspot...rch?q=gymnasts