You're pointing to a much bigger problem than one that is covid specific, and that's the trouble with retrospective studies generally, and small retrospective studies in particular. Retrospective studies area great for a singular purpose, which is to find interesting questions to test on a prospective study. Being that they are close to worthless. And yet...everyone uses them to make a point. Typically this is confined to advocacy groups trying to advance an agenda. However, lately it's gotten really bad because of the stakes and the politics. Every loves was quick to point out how shitty the Henry Ford hydroxychloroquine study was (and they were correct), but apparently just because the president was satisfied. When it comes to making a point we agree with however, the ends clearly justify the means. Science is hard. Data interpretation is hard. They're are no easy answers here, so we need to always keep an open mind about what the data say, and not what our favorite or least favorite politicians or media sources say. Even a blind squirrel can find a nut sometimes.