I advocate for this mitigation: moderators approving the first post of an account.
Currently, the spam gets removed manually and quickly, therefore the first posts would get approved at a similar rate. It's privacy focused and will stop anything but extremely focused human engineered spam.
It's like avoiding badness enumeration in a sense. It's impossible to list all bad actors, so why not only allow good actors. Of course it can be combined with any other methods that are desirable, though.
I can't speak to whether the burden of manually approving every forum user would be an unreasonable ask of the moderation team, but my idea is only the first post of a new account, which is not many and could be quickly eyeballed in a unified interface. It seems to me that it would be easier than hunting for specific first-posts that are rule-breaking.
Many forums like Privacy guides.net, monero.forum use this technique.
I don't know if it's possible on Flarum so that could easily disqualify this approach.
Curious to hear the admins, moderators or anyone's thoughts on this proposal.