Unless I've misread something, that's not what the paper calls for at all. In fact the paper seems particularly concerned with avoiding excessive centralization of gatekeeping powers.

Decentralizing gatekeeping power is the opposite of censorship.