If you understood where I’m coming from you wouldn’t rely on the academic origins of these ideas to convince people to trust them. 
Skimming the discussion you linked about unions it seems that on the one hand, there’s clearly something wrong with how it’s formulated in PEP 483 (and we should try to come up with a PR that fixes it), on the other hand it feels like a bunch of people yelling at each other “you’re wrong” – “no, you are wrong, you misunderstand me” – "no, you misunderstand me. 
Going back to the subject line of this thread, it feels like whatever the benefits of a set-theoretic type system, “less divisive” does not appear particularly likely at this point.
I’m happy to let it rest – we should probably focus on smaller problems where we can get agreement. Gradually improving the specs sounds like it will be useful, but it’s not the only thing. I’ll see you in the discussion about the Typing Council (eventually, I have other things to catch up on).