Most of the meat is in the second half of the video. The interviews and case studies are precious records of frontline experiences, but it is also clear that both the sampling and the analysis are biased.

The sampling is clearly biased towards instances more friendly to minority groups. The limited visibility due to the network structure and the moderation tools is beneficial for privacy and safety, but I fear it also makes sampling across the federation gaps harder.

The analysis is clearly biased against centralized SNS providers, with barely any credit to their achievements. It nods to the positive correlation between the moderator-user ratio and the difficulty of moderation, but did not develop it to its logical destination. If 1 moderator to 9000 users is already resulting in a "black hole for reports" situation, what is the ratio for Twitter? How do they manage to keep up?

The case study of Gab is a good showcase of available moderation tools, but goes barely any deeper. The central phenomenon, "islands", is mentioned only in passing as a precursor to Gab's defederation.

The study is important work, but it is also clear that more work is needed to understand where the Fediverse is situated in the SNS concept space, and where we may want to take it.