change hivemind healtcheck to start after 3 days to ensure haf has replayed
We started caring about this healtcheck too soon (5 minutes), so if haf hasn't finished syncing, hivemind block processing would error out, killing docker compose. Changed to 3 days.