I believe the author of this piece is keenly aware of the issues inherent to study designs such as this, which is why he discusses the methodological issues in

I believe the author of this piece is keenly aware of the issues inherent to study designs such as this, which is why he discusses the methodological issues in this post. He has discussed methodological issues at length in previous posts (he is a psychiatry resident, and one who frequently goes back and addresses the primary literature) and that is taken as context to his regular readers, explaining why he doesn't devote more time to disclaimers. Your exhortation that he should spend more time on basic understanding of statistics is misplaced, because he understands these issues, but he has different philosophical views than you on how information should be gathered and processed. Specifically, it seems to be your view that information received with noise and bias should be given (literally) zero weight, that information should be ignored unless it comes in the form of peer-reviewed, random controlled clinical trials. His philosophical view is that all information should be given some appropriate weight, and that the costs to acquire information are relevant. Random, controlled clinical trials are costly, and so we will never have them on a variety of topics of interest to us. To account for this finiteness of resources, and to suggest which areas could be more fruitful for future formal trials, it seems wise to acquire that information which can be cheaply had. This is an attempt to gather that information.