Someone deleted the rows at 14:00 and the backup is green
At 14:00 a maintenance script ran a `DELETE` without the intended `WHERE` clause against a 1.2 TB production database. It was noticed at 14:20. Nightly `pg_dump` backups have succeeded for 400 consecutive days, into a bucket in the same cloud account. The stated RTO is one hour. Nobody has restored this database since it was 40 GB.
The fix that looks right
Restoring the nightly dump directly over the production database, because it is the documented procedure and it is the fastest path to a database that contains the deleted rows. It does restore them — along with reverting every order, payment, signup and update from 00:00 to 14:20, silently, with no record of what was destroyed and no way to recover it, because the current state was overwritten by the restore. The delete affected one table; the restore affects all of them. A logical deletion is almost never a reason to replace an entire database, and the moment the restore begins, the option to recover the rest of the day is gone.