I'm not convinced the reasons he gives for why we're so good at developing athletes are the actual reasons; I think they're methods we do it by, not the reason

I'm not convinced the reasons he gives for why we're so good at developing athletes are the actual reasons; I think they're methods we do it by, not the reason for which we do it. The real reason we're good at generating athletes, is that they generate large amounts of revenue for the companies they work for. The entire system for creating athletes (youth opportunity, publicly celebrating their success, etc) exists because they make huge sums of cash for their eventual employers. It's also easier to compare athletes. Each sport has a canonical set of statistics kept. I can easily compare athletes #1 and #2, and decide which is worth more money based on their stats. Not so easy to do with writers, so much of it is subjective. There also exists a quick and easy test for determining athleticism: the game. They're short, and they generate data. You can hold enough of them in a short enough amount of time to get the data you need to make decisions about/develop an athlete. If we're talking about them as products, athletes are much easier to iterate than authors; the feedback loop is much shorter.