Implement a dynamic etcd revision bump calculation based on the last known maxRaftIndex (from revision.json), using the static 1 billion as a safe fallback. This change is wanted to optimize etcd resource usage; a proportionally calculated bump is preferred over the large static jump because it consumes only the necessary amount of the int64 revision space, thereby extending the overall theoretical lifetime of the etcd cluster.

This change would also align podman-etcd to what cluster-etcd-operator quorum-restore-pod utility does already.