The problem is that all those bad decisions inevitably spark from these platform companies not wanting to enforce their own AUP.

We point out a particular bad actor in violation of their own policies, and they do nothing. We complain loudly about it in the news, and they vaguely commit to "fighting" the problems that were already against their AUP... but not any specific named bad actors. Automated systems are chucked at the wall, but the actual people people are angry about are exempted in the name of "free speech" until it costs the company business.

This cycle repeats until a politician gets a bug up their butt about it and passes a law mandating platforms have more censorious policies.

The fact that Clouflare has decided it is in the better interest of the Internet that one website gets to break their AUP because they've become a cause célébre is so amateurish. Business to business relationships don't work that way. You break the contract, you're out. That's what's in the interest of the Internet. Upholding business contracts.

Reminds me of Twitter's world leaders policy they trotted out around 2017 to keep Trump on the platform. If I remember correctly Trump was even immune to DMCA 512 takedown notices, which is an absurd level of risk to take just to say you have @POTUS.

It's also sort of hard put the cat in the bag, for the service that instituted the policy and for competitors.

If you start a competing free-speech twitter competitor, you're effectively "twitter for InfoWars fans, QAnon and assorted holocaust deniers". Likewise, the headlines if Twitter themselves softened their line would hardly be about standing up for free speech ideals...