If the solution was providing people with a list of bad faith tactics we would have been done with it at least twice by now: first when Socrates was arguing with Sophists 2500 years ago and another time when Schopenhauer wrote Eristic https://en.wikipedia.org/wiki/The_Art_of_Being_Right. And before you think 'maybe someone doesn't know yet': yes, you are correct. Someone doesn't know. We tried telling everyone before and just trying harder doesn't seem to cut it.
What, I'd argue, would be at least a tiny step forward would be thinking in terms of games people play, rewards they seek and maybe even monetization of systems. Thinking of people who argue in 'bad faith' as being mostly plain wrong is naive and somewhat offensive. Talk to your PR department every now and then, some of them are smart and know what they're doing. Same for Twitter discourse and all else.
Telling people (or yourself) to make better communities ignores costs involved in managing that community. Can you afford onboarding of even telling people that cute list of bad faith tactics? Can you do it faster than a place that doesn't do it? Can you achieve retention higher than love bombarding communities?
No. No, you can't.
Not with current tooling at least. Not to push own products/services (today!), here are some angles that seem achievably hard, yet somewhat underdeveloped: good faith arguments are more time expensive - it can be cut into pieces/redesigned to give them more chance; both wrong and correct ways of thinking about specific problems are actually very limited in numbers - maintaining searchable database of them to reuse should dramatically speed up 'getting through'; false positives in ostracism are unnoticed - layered moderation that provides feedback on initial misjudgment can noticeably improve the space: not so much retention (that numbers would be small), but limit echo chamber by avoiding rituals of cancellation - without increasing costs as much as having 'full conversation' with everyone before banning would.