While I'm all for debates about the pros and cons of more or less permissive speech policies, sentences like this always betray the quasi religious nature of American speech attitudes. Many of the oldest democracies in the world work perfectly fine and do not follow this line of thinking.
It would probably be worthwhile to have a much more empirical approach, trying to actually measure what the impact of prohibiting certain speech is. How many people are unduly silenced, how much intended harm is prevented, and so on. Because statements like:
"It is quite clear that the perceived benefits of censoring psychically harmful hate speech are far outweighed by the costs of such suppression."
aren't clear at all. We often don't know because these are really not evidence driven claims. In our information age there should be a rigorous science on how moderation of speech impacts networks and it deserves fairly high priority.