It would get deep into the grey area if Google users had any capacity of enabling the communications with those sites. As of today, Google is the one in control of the communications, and dictate who can reach anybody over most of the internet. They just don't have a policy of empowering their users to decide who they want to talk to.

What speech is being censored is harder to discover. They are blocking people from communication without any feedback, and it would take a large effort to reach them and discover who they are. Certainly most of what is blocked is spam, but that's true for whatever block you implement today, unless you spend an unreasonable amount of resources targeting it into non-spam.