What would we prefer, a study that relies on self-reporting but is a study nonetheless, or a bunch of theories based on entirely on anecdotes with a sample size

What would we prefer, a study that relies on self-reporting but is a study nonetheless, or a bunch of theories based on entirely on anecdotes with a sample size of 1? I get this study isn't perfect but it seems silly that so many commenters here are disregarding it entirely yet are perfectly happy to listen to completely untested theories from people with no expertise whatsoever in the field.