Moderation becomes much easier in a decentralised, federated platform. Anyone can post anything, but community standards become more localised by virtue of being instances of a larger system.
In any scenario, though, the focus has to shift from problematic individual comments to problematic individual users and/or accounts, and there needs to be investment in both a moderation system (user-facing and back-end) and moderators who can identify them. Facebook and Twitter have shown no serious inclination to even attempt this, and instead we have situations like the one described in the article.
