Currently, the logic to wait for a new config to become committed after a safe reconfig lives inside the processReplSetReconfig method of ReplicationCoordinator. We should move this logic into its own, public method on ReplicationCoordinator that external clients can call separately after calling processReplSetReconfig. This will make it easier to write unit tests that don't want to block the calling thread to wait for a config to become committed. Since this waiting logic isn't required for enforcing the safety of safe reconfigs, it is acceptable to move this to a separate public function that can be called, or bypassed, by an external client. For the reconfig command, we will add a call to this new wait method after we process the reconfig command.