RDS manages database infrastructure, but you still choose the recovery and access design. A Multi-AZ standby supports availability and failover. A read replica supports a read-scaling or replication use case and can have replication lag. These are different roles, even when both involve another database copy.
Backups support restoration to a recovery point. Restoration isn't the same event as automatic failover, and you must test the application against the recovered endpoint and data. Keep private databases private unless a specific requirement justifies public accessibility. In the exercises, inspect engine, availability, and backup settings independently. The edge case is a system with a replica but no accepted failover or recovery plan: another copy alone doesn't prove the required recovery time or data-loss target.