Read replicas and stale reads

Almost certainly. Route reads that follow a write by the same user to the primary for a short window, or track the write LSN in the session and wait for the replica to reach it. The general fix is read-your-writes consistency, not lower lag.

Deleting rows marks them dead but leaves index entries until vacuum reclaims the pages, and even then the pages stay allocated to that index. REINDEX INDEX CONCURRENTLY is what actually returns the space.

Doing it once and never verifying. The setup is the visible part, so it gets attention, and the check that would catch a silent failure never gets written.