>The argument being argued here is that it's not viable in practice for Facebook to police every single post, not that it shouldn't do it in principle. Your argument seems to be that Facebook shouldn't police messages at all
I'm pretty unconvinced that they should in principle be able to both engage in content moderation and retain safe harbor.
Cellular providers do try to block spam and certain kinds of fraudulent/impersonated content.
A shopping mall isn't liable (in most cases) for the products or services sold by the individual stores yet it can choose to not allow certain businesses (tobacco, gambling, 18+ stores).
Have you seen a completely unmoderated social media site? It's either ultra niche or riddled with spam, fraud, and other low-quality content.
It's really simplistic to only bring up social media when the larger topic is "should private control of common platforms, markets, and network effect relationships be allowed"? The real estate and many critical B2B industries are FAR more insular and controlled by arbitrary groups.
Should McDonalds Inc. be liable for franchises selling expired food and/or should they not be able to control a franchise at all? Franchise contracts are way more controlling than any social media ToS yet people pay to join them.
I'm pretty unconvinced that they should in principle be able to both engage in content moderation and retain safe harbor.