His argument is that it's inherently hard to do exercise science research because of selection biases and sample sizes. This is a fairly reasonable statement.

The problem is confounded because of conflicts of interest. Either the investigators are being remunerated by a commercial interest or the investigators are hoping to establish a commercial entity.

It's not uncommon for a study to come out with interesting results and conclusion only to be found to be completely unreproducible. It takes a while for this to happen, and even then, people are confused over who is right.

Also, linking to the wiki page for a logical fallacy is a very passive aggressive swipe and low quality.