(just a few thoughts I wanted to write out)
Don't get me wrong, I love the local and federated timelines, but after thinking about it I realized that it's also the cause for a lot of drama.
Email and Xmpp never had such a big problem with cross instance blocks. If you think about it, all federated content is blocked by default and only becomes available if a user searches for it and subscribes to it. Before that, the server has no idea what is out there unless a relay is used. But there's two exceptions... the local timeline and the federated timeline.
These are great to get stuff started and kickstart the following process, but are forcing people to receive content that they might not want to see.
Where previously a block would only be necessary whenever a malicious user messaged me directly, now we have to deal with the need to curate content of public timelines in order to avoid problems with local or remote users.
The instance admins have full right to decide what is hosted on their instance and what not. This is not about free speech because you are not entitled to using someone's server in a way they don't want, but about creating complicated dilemmas and tough moderation choices by forcing together content and users that could be drastically different in beliefs or preferences by using timelines which are understandably very appealing to use.
Maybe all posts should be unlisted by default and both timelines, whether on Lemmy or Mastodon only contain whitelisted user accounts to give your instance's users and remote users a few recommendations.
Don't get me wrong, I love those two timelines and I have a thick enough skin that I can simply ignore or block content I don't like, but as an instance admin both on Mastodon and Lemmy I've noticed that this is not the case. Users are often eager to report anything they don't like to see or disagree with even though they don't follow that community in question or would never have interacted with it. This could cause a lot of moderation overhead as well as drama as it puts users and remote instance admins on alert about X, Y or Z divisive or distasteful content (especially when it comes to NSFW) potentially being sent out to them.
Yes, that might actually be the activitypub table in the database. You can safely delete older entries, like from two weeks ago or older. Otherwise it just keeps growing with the logs of all activitypub objects that the server has sent out received.