m4v3rick Flarum scales really badly for large discussions, because when you visit a discussion, it fetches ids of all posts in this discussion. If you have discussion with 10k posts, that is 10k posts that needs to be fetched from database (AFAIK only ID and type, but still this is 10k rows to find in DB), processed and generate JSON for them. And then client needs to work on this massive array of IDs.
Discourse has the same design, thus the same problem, and they decided to just not support this use case: https://meta.discourse.org/t/the-megatopic-public-good-or-public-menace/85592