Honestly, parsing through version history is actually something an LLM could handle. It might even make more sense of it than without. For example, if someone replies to a comment and then the parent is edited to say something different. No one will have to waste their time filtering anything.
Doesn't matter. If he joins he was lying. If he doesn't join, it doesn't mean he was telling the truth. Nothing he says matters and if he does the opposite of what he says there's no action to take.
Didn't they use to have this years ago? And then removed it?