Restoring from a backup could be a time consuming operation due to Rollback To Stable which is required as checkpoints are fuzzy.

By having the ability to take stable checkpoints instead of fuzzy checkpoints could greatly improve the restore/recovery times as RTS will become a no-op.

Backups are taken from standby/secondary nodes. With the ability to configure the checkpoints as stable, the server can configure WT to take stable checkpoints when a node is designated as a secondary/standby node. Subsequently the backups taken from the secondary node will be based on stable checkpoints and the restore from such backups will result in RTS being a no-op and can reduce the recovery time.