The trouble with the tech giants is that their "free service paid by ads" business model completely collapses into red numbers if they start employing adequate numbers of people for that purpose.
So they resort to artificial intelligence, which is worse than natural stupidity in this regard.
If that existed, I'd refuse to be part of whatever community you think I belong to and I'm going to pick the one least likely to interfere with me.
And if it's a forced choice, I'm going to fight that instead. I think you'll have to educate and then just let everyone say whatever they want.
You know, you don't HAVE to remove a video. Can just put another video above it that says "here's what we think is going on" and be done with it.
For example, Czech Facebook banned an ad with a word "Rifle" in it, because of an American ban on gun advertisements. But "Rifle" means "Blue Jeans" in Czech and nothing else. It is not a gun-related word in our language. And indeed the ad tried to sell blue jeans.
Consequently, you get someone making minimum wage banning videos because they're not saying only positive things about vaccines.
And Google and Facebook won't care. Because moderation is a cost intended to curb the worst PR scandals, but content is profit.
If it's defined by social consensus then during periods of groupthink and hysteria (common among humans) even the most reasonable people will be labeled wackos and shunned.
This fits in line with a common conception of what we view as reasonable limits on free speech.
Removing speech of people sharing their beliefs about the risks and rewards of putting substances in their bodies or other medical decisions doesn't meet this standard.
That's different from spreading conspiracy theories about putting substance in their bodies. Actual disinformation (blatantly false) and not just discussing their hesitancy to get a vaccine.
Content moderation is not inherently good or bad, it simply is. A version of Youtube that cracked down on well known science would be legal but not "good".