I don't understand. Isn't this already the case in America with Section 230? (The original post is about the Netherlands, so this is now tangenting.) It's just that no one actually acts as a platform.
I'll daydream with you except mine is different. The optimal social media in my view is one tied to your real identity. Moderation would only be applied under court order by the relevant jurisdiction for the view of the content within that jurisdiction. i.e.:
1) American posts content critical of Indian officials. That content is restricted by order of an Indian court and no such order is additionally given by an American court. It would be hidden from view within India but not from within America. The inverse would be true.
2) Indian posts content critical of Indian officials. That content is restricted by order of an Indian court. America (or any other nation) has no duty to protect that speech and thus no claim over it. That content is censored everywhere.
Additionally, everyone would have client-side filters which may be published. Emphasis on "published" because the publisher would be accountable for their words just as much as a newspaper within their jurisdiction. Though they wouldn't need to say much (i.e.: list of people I [dis]like). Unique identity and nationality are the only ones I can think of right now. More complex examples:
1) An American publishes a list of politicians who have made inflammatory public statements. They have evidence of this for each person on the list and make no additional assertions about their behavior. People not interested in such content could subscribe to the filter. (I guess people interested in chaos could view the inverse.) No court is willing to censor this list because their statements are protected speech.
2) An American publishes a list of men who have committed sexual crimes (such as in the original post). They assert it as fact not alleged crimes. They include someone who has not been proven in a court of law to have committed that crime. They can be sued for libel and possibly forced to remove the person from the list or reword the list description.
Anonymity between the user and the social media service wouldn't exist, but it might between users. The service could be mandated by the jurisdiction to unmask or otherwise ensure the accused does not fall within the jurisdiction.