Though with international platforms jurisdiction with law enforcement can be tricky, I'm sure I remember cases brought to court successfully off the back of Twitter abuse.
Not disagreeing with your points entirely though. Perhaps there is middle ground in the platform providing analytics of this kind to law enforcement?
On one level there's the immediate individual occurence of harassment where I agree law enforcement should step in where applicable, instead of putting that responsibility on the platform owner (though it doesn't seem like law enforcement is typically equipped or motivated to do so, dedicated analytics seem like a good step towards addressing that). But I don't think that needs to preclude the platform taking action as well, simply to improve their product for the vast, vast majority of users who don't get off on harassing marginalized people.
The other level is the macro scale effect of being the platform that's cool with white supremacists chasing other people away. Independent of your motivations for having a hands-off approach to moderation, you're gonna be emboldening them, making them feel welcome and comfortable and consequently growing a toxic user base. In turn, people who are particularly subjected or vulnerable to that sort of harassment will act in their self-interest and get the fuck out. "No moderation" isn't a neutral option, and law enforcement isn't going to take care of community management for you, so I think this is a strong argument for the responsibility of the platform to step in.
I'm not sure if the royal mail comparison is really applicable. I guess you can frame every platform as some sort of public utility, but I think that analogy works better for lower-level sorta things like OS developers or CDNs. In particular I think the argument doesn't do justice to the role of twitter as a social space. Maybe a better comparison would be whether you'd expect a bar to continue to provide service to patrons with a history of starting fights with other customers? You'd also expect law enforcement to arrest people for assault, but I think it'd be reasonable that they wouldn't be let back in after trashing the place anyway.
What would you say to banning and retaining the history. Tweets all appended with "user banned for x"? Like a bar with photos of banned patrons.
But there's a lot of nuance between hiding them, and not giving them a platform. The points of view of the President-Elect aren't in any danger of disappearing from view, they'll be published and discussed in the media regardless, and there's been plenty of white-nationalist websites with discussion forums for everybody in the alt-right to present their message.
Another area where one might find nuance is between expressing viewpoints, and harassment and hate speech. It's perfectly possible to express oneself and one's politics respectfully. I don't think that's what anyone is talking about when there's talk about twitter needing to clean up their platform. Contrasting viewpoints are not the point of contention, it's whether people should be supported when they go out of their way to hurt and abuse and intimidate other human beings.
Should twitter really feel obligated, morally or on any level, to sacrifice the well-being of the victims of slurs and harassment, only because of the (very tenuous and abstract, in this case) danger of censorship?