Polling at theaters done by one or two guys is not a reliable data. Keep in mind that someone who did podcast may cherry pick the results, and what is more important, asking dozen or several dozen people in few theaters what they think doesn't mean that "most black people" think the same. Maybe I am wrong and it was big, reliable poll, but it's hard to believe that some podcaster did reliable poll, with unbiased questions and proper group of responders.

Usually polling done by people who record things for youtube/podcast etc could not be considered as reliable data, only something humorous.

If this is a basis for assertion that "most black people do/think X" then I believe that you could find a basis for any assertion like "most X people do/think Y". I think it's bad to generalize things like that based on some podcaster polling.