I think Sabine explained this social effect few years ago. I know she's a little controversial, but the key thing in the video (as opposed to all other videos about DSE on the internet) was that you don't get "two clusters" actually. They are both statistical parts of a single [non-]interference pattern. "||" is a lie. I'm not in a physics rebel camp and don't prefer Sabine either, but after that I sort of lost trust in the interpretations that can't even get the resulting picture right. I even suspect that showing dumbed results amplifies "wow" effect and monetizes better.

https://www.youtube.com/watch?v=RQv5CVELG3U

This is the video if you're interested. Again, I'm no physicist and don't know if explanations are legit or statistically correct. But that little || trick that all other popsci videos play on you, that's a true concern.