I don't think this is the case, so I'm led to think there's some other flaw in this line of reasoning. Notably,

> Censure was once only available to the government, since they ostensibly controlled all public spaces and had the forces to make it so.

Is plainly false. You've never had the "right to distribution". A printing press owner didn't have to publish your pamphlet. The church never needed to let you give a sermon. The NYT didn't have to publish your article. CNN has never had an obligation to let you on air. There has always been a "best" method of distribution, and its always been private.

Twitter blocking Trump isn't functionally different than the NYT or WaPo publishing the pentagon papers. It is a private organization exercising their right to push back against government propaganda. It is first amendment protected, and it should unequivocally be defended by those who believe in civil liberties.

If it's based on average users over a time period, do we trust the site to report it accurately? If a user gets banned from a "private" site that has not yet hit the size threshold, and the user complains to the government and alleges that the site actually is big enough, does the the government go in and audit the company's logs to make a determination over whether it should be a "public" site?

If it's based on something looser, like "percentage of people who use this site as their only news source," (which does seem more defensible, to me) how do we set up the criteria for making the determination?

Also, how would the federal rules for content moderation look? Would a call for open insurrection still be bannable? What if two roughtly equal portions of the electorate don't agree about whether it was a call for insurrection? Who decides if it was?

I think questions like these need to be thought through and answered before we can seriously talk about making a law that prevents social media companies from banning anyone they choose.