Generally the recommendation is to manual fail back after a failure event. The reason is for someone to verify the issue causing failover in the first instance is resolved and the system is stable.
The danger is that if there is an ongoing or intermittent issue and you have revert configured with preempt you can end up with a flapping of failovers back and forth continuously until the issue is resolved.
If you do configure preempt be sure to select a long enough period of time to be reasonably confident any issue has indeed cleared to prevent such flapping.