A hypotheses formed over beers but never tested outside the bar is like most other beer-centered conversation: of limited value. Science communication, similar

A hypotheses formed over beers but never tested outside the bar is like most other beer-centered conversation: of limited value. Science communication, similarly, can start anywhere, but what you're describing has a very tenuous relationship with people talking on YouTube. I'm all for people doing science having as much flexibility as they want, up to the point where what they're doing is no longer science. People forming ex post facto hypotheses based on carefully-chosen data, or bypassing any peer-review in favor of YouTube communication, that should raise serious alarms bells for anyone in the sciences.