There's some good stuff in there and it's easy to cheer for some big new regulatory burdens being put on Google and Facebook, but it's slightly chilling to think what it'd be like if they eventually try to apply it to the fediverse. It sets up teams of what it calls "trusted flaggers" for example, whose job it will be to scour the net for anything they believe to be "illegal content" and order it removed. I imagine they'd start with places like c/piracy, but once such a vast apparatus for net censorship is set up who knows where else it might start looking. They'll use it to go after sellers of "counterfeit" goods as well. Imagine your instance admins being forced to go through some kind of appeals process to take down posts they don't like, but being required to instantly take down posts the government doesn't like.
I don't know, it's pretty complicated but there are some reasons to be slightly worried about it I guess.