At a certain point, the welfare of society outweighs the freedom of the individual. And you agree with me. Hence the phrase "nearly unlimited".
That's one of the supporting principles of all totalitarian regimes. The problem is who defines what's the "welfare of society" (Normally is "the Party" or "the Leader").
Individual freedom is an inalienable right, and should not be curtailed, unless you're physically hurting others or infringing on their own freedom.
How about if you're planning to hurt others? That's "conspiracy" in most "totalitarian regimes" (that is, societies with laws). Is it a crime to point a loaded gun at someone, or only to pull the trigger?
How about incitement? Also a crime for any government. If someone encourages a murder, are they at least partly responsible for the murder?
Your definitions of freedom and oppression here are trivial mental exercises that fail in real-world application. Who defines what's the "welfare of society"? Depends on the society. You're confusing who defines it with how it is enforced. Governments have a monopoly on violence in the ideal case. Who controls the government? If it's a representative democracy, then it's more or less the will of the people.
I've seen this bullshit argument before, rather a lot. I saw it made when I asked if it was okay for the neo-Nazis in Charlottesville to surround a synagogue, torches in hand (which they did). And people with similarly shallow, trivialized senses of freedom and society argued that such behavior was fine, as long as they stayed on the public sidewalk.
Is enabling fascism a form of fascism? That's a good question for the theorists. As a friend of mine put it, "If you have to ask what the definition of fascism is in order to argue that you're not a fascist, you're a fascist".
The rule of thumb that I go by is "we should allow whatever it is that the United States Supreme Court allows".
The thorny edge cases that you are talking about have been explored in depth by the Supreme Court, and I think that they have made good decisions.
(IE, the good decisions that they have made is to almost always side in favor of free speech).
Further The USA SC is an unrealistic expectation and pointless bar of value when talking about Facebook and international speech.
I am not talking about the time that it would take to decide a case. I am instead using the Supreme Court as a general example of where on the scale of freedom/ censorship that platforms should take.
The location on that scale should be "What would the supreme court do?" I think that they have chosen a good balance, is why.
Actually that's one of the supporting principles of all regimes, period.
If YouTube's algorithm recommends a hateful propaganda video to me, I shouldn't be angry at Pepsi for the preroll ad, they obviously don't specifically endorse that content. But the acceptance of targeted ads implies that endorsement. I just wish we could entirely abolish targeted advertising, it would make the internet overall a better place. And it wouldn't even make the ads less relevant, considering that Amazon and Google apparently can't recommend products I'm interested in even with all the data they've collected.
We essentially need a Google (the framework which they use to algorithmically understand web content) for social networks with a front-end UI for topical mapping.
--
Also to add to your idea, I'd be very interested in a social network that can do that, but also requires real ID's and has some sort of public display of the topics that any user talks about and engages with often.
That way if I'm talking to someone I can gauge if they're really someone I want to be talking to about a particular topic. (can also be used to better map taste to reviews of products/services/companies/etc)