It is recommended that restoring a cluster must either:
1) Restore to a new Atlas cluster and reconfigure your application to use that new cluster once the new deployment is running, or,
2) Ensure that the target Atlas cluster cannot receive client requests while you restore data.
(src https://www.mongodb.com/docs/atlas/backup/cloud-backup/restore/#prerequisites)

While doing a PIT restore with the source and target cluster being the same, our clusters have gone into a "down" state 3 times due to obscure errors, opslog not being large enough, or writes happening to the cluster during restore.

We would like to take the safer route of always restoring to a new cluster.

However, in doing so we lose the Atlas snapshots and possibly opslog data.

Feature ask: Ability to copy or move snapshots from a source cluster to a target cluster. This operation would take place after the primary restore completes, if needed.

A similar feature was asked here, but not exactly the same thing: https://feedback.mongodb.com/forums/924145-atlas/suggestions/44074839-restore-snapshots-on-s3-back-to-atlas-cluster