After what seems like a thousand years I have started replacing some synchronous client to server #ActivityPub fetch requests with loading the same information from an indexed cache in my link aggregator.
I think I've started the work on inbox/outbox async fetch and indexing around 3-4 months ago, and only now I am able to make proper use of it.
I guess I should have used a real RDBMS instead of reinventing the wheel on every turn. :D