> If you're going to murder your parents over a conspiracy theory is there any reason to assume you wouldn't murder them for a dozen other reasons?
I'd wager that some will and some won't, and that if you were morbid enough you could probably measure such a thing statistically.
> How justified is censoring a platform for millions of users because of one mentally ill user?
Not only the mentally ill occasionally have absolutely terrible ideas, buy into something stupid, or get occasionally conned. Adding information to videos is also not censorship - at least insofar as I'd define the term, or generally seen it defined, unless perhaps if you're going so far as to try to bury a needle in a haystack - so I'm not sure the question is relevant here unless you're worried about some slippery slope or something.
I'd hesitate to impose censorship, but adding information to pierce the walls of echo chambers - and to let e.g. "the marketplace of ideas" actually work - seems like a great idea to me. Not that any implementation of such a thing won't have it's flaws, mind you.
> Is there any evidence that being exposed to certain ideas "pushes someone over the edge" any more than what they can come up with in their own head or without technological assistance? I've read some stuff about school shootings that suggests media coverage encourages "copycat" attacks.
I feel like you've answered your own question here, but I may be missing something. Isn't what a copycat being exposed to when they read the news about one of these incidents, also ideas? Terrible, terrible ideas? Or are you perhaps asking if any of that stuff you've read rises to the level of evidence instead of clickbait hearsay? I've likely read little more than you have - references to studies that suggest the evidence is there. I don't recall reading about any studies refuting that idea or showing no correlation.