We already have user liability under the current interpretation of 230, and certainly I agree it should come with strings. The problem is that there are perfectly legitimate reasons to for sites to remove posts. For example, it would be reasonable for Stack Exchange, a developer oriented site, to remove legal and inoffensive posts that are off-topic (e.g. cooking, football, 70s rock). Their purpose is stated very clearly. Similarly, a site like thedonald is absolutely clear in their purpose, so it would be reasonable for them to remove anything critical of Trump - just like the same would be true of a Biden fan-site. Any reworking on 230 would need to prevent de facto publishers using 230 as a shield while still allowing platforms to manage content in a way that reflects their stated purpose. And none of this vague 'hate speech' nonsense to excuse removing people who criticise migration while allowing overtly racist left-wing content.

It's people smarter than me who'd come-up with a way to make this work.