Allow to store all data on s3 (blob and metadata)

## Is your feature request related to a problem? Please describe.
Currently only the blobs are stored on s3 with s3ng, the metadata are kept on the local pv. This makes a scaled and distributed installation unnecessary hard because there is a separate distributed / replicated storage needed. Modern s3 stacks like minio replicate the data on their own, so there is normally no need for a complicated distributed storage like ceph, longhorn, glusterfs or others.

## Describe the solution you'd like
When specifying the s3 storage driver I would expect that all data (Blobs and Metadata) are stored on s3, so that a additional pv is not needed anymore.

## Describe alternatives you've considered
Well, not really one, because the need to have a distributed filesystem does not really exist anymore in our environment. Blob storages replicates themself and databases also. The configs normally reside in configmaps and therefore are replicated by k8s.

## Additional context
I would expect that this should be possible, because high data apps like Grafana Loki (Logs) or Grafana Mimir (Metrics) are already able to handle this.
I know that the old s3 driver stored all data on s3, but not in an efficient way. But I think there should be a solution, because this would really help to reduce the administrative efforts (backup, deployment, ...)