This take is inaccurate and a myopic understanding of free speech.
Moderating social media, and the internet is in fact, doable, and absolutely necessary. This has nothing to do with "censorship".
The debate is actually easily settled if you understand what is happening with free speech online.
What typically happens in conspiracy-circles, is that people are radicalized because the disinformation is simply not challenged. It may be that a few users will dispute various claims, but their valuable, fact-based input, is typically drowned in a flood of spam, personal attacks, and claims unrelated to the claims that are being discussed in a given forum- or comment thread.
The problem with "unmoderated free speech" is that informationterrorists can abuse "free speech" to repeat the same disputed claims over and over, without ever addressing the fact that their claims have been disproven. This is also what I would label as "flooding the discussion" or "drowning the facts"; it is so effective that everyone who conducts themselves properly and respectfully are drowned in this flood of disinformation; this actually results in a "suppression" of free speech. When only one side is really heard, then we effectively do not have free speech.
Instead, what we have is a conversation that is dominated and suppressed by a few bullies that are shouting the loudest.
In addition, you would really hate to have governments influence the fact-checking processes on social media platforms, since governments have ultimate power, they are also the largest threat to free speech. Ideally fact-checking should be done 100% transparently by independent fact-checkers, and the facts that lead to a conclusion has to be tediously and transparently documented so that everyone can trust the processes. People who think the conclusion of a fact-check is inaccurate should take it up with the relevant fact-checkers, or possibly take it through the courts.
This "ideal" of "unmoderated free speech" has never really worked. It did not work in the real world, and surely will not work on the internet. The problem with this idea is that anti-social individuals will just try to control the narrative by spamming or repeating disproven claims (shouting), making new false claims, pushing disproven conspiracy theories. Etc.
A common technique I see used by malicious sources, is to release one claim, have people debate- and disprove it, only to release another, unrelated, claim without ever acknowledging the fact that their first claim was false. The result is that even old and disproven claims are circulating in an endless loop. They use this technique continuously with countless of subjects, both old and new — you would think that people will eventually reject claims made by known informationterrorists, due to their lack of credibility and history of publishing falsehoods, but that does not seem to be the case.
I am not a fan of banning people permanently from social media, as it just seems too merciless — there has to be ways to get un-banned — but, as a minimum, we should have fact-checking on profiles with large followings; and of course, groups and profiles used primarily to spread disinformation should be deleted.