Well, the case of incitement to violence (willful or no, meeting the legal definition or no) is a tricky one. With almost every other case I would say no, the block might be justified by policy, but the policy is ill-advised inasmuch as Twitter is a conduit of public discourse.

In cases of imminent harm, there is a very good argument that public discourse is not impeded meaningfully by a ban or delay, so the upside (avoiding sizable danger of loss of life or damage to property) outweighs the downside (you'll have to wait a while to use this particular channel to interrogate the validity of those bad ideas).

Then the question becomes: if we accept that imminent harm is cause for temporary abrogation of free speech (much like a curfew is a reasonable response to a riot), how long does the danger period of risk of imminent harm last? That's hard to say, but probably not 16 months.

Finally, the last, most important, and most difficult question is: does Twitter as it stands actually serve as a conduit that has the requisite properties to gain the benefits of free speech?

Right now, my answer to that is a resounding NO. This means that if I'm answering a question about actual Twitter as it stands right now, I'm not going to give the free speech answer, because Twitter badly fails the preconditions necessary for the arguments in favor of free speech to work. In particular, its mechanics exacerbate factionalism and enrich for falsehoods while minimizing opportunities for ideas to be tested against each other.

So--remove free speech in case of imminent incitement: yes.

Remove speech for months and months: no.

Free speech concerns are relevant to Twitter banning people: not really (not for the lame "private company" reasons, but for important structural reasons).

I know how to fix Twitter: no.