Currently, `containers/storage` includes functionality to convert container images that are not in `zstd:chunked` format to `zstd:chunked` on the fly so they can be used with`composefs`.  There is no technical requirement for this, but it is done this way to simplify the implementation and reuse the existing code path:

https://github.com/containers/storage/blob/78f4258b2bd9abb338a407e6799cf99565eabb10/pkg/chunked/storage_linux.go#L1413-L1478

For images not originally compressed with zstd:chunked, the current implementation compresses the tarball using zstd and then immediately decompresses it to prepare the data for content-addressed storage and composefs processing.  This is unnecessary.

To optimize this workflow, we should introduce a "dummy compression" option or logic within the conversion process. This dummy compression would act as a passthrough, allowing the data to flow through the existing `zstd:chunked` conversion code path without performing actual compression or decompression and work directly on the tar stream.

This is useful to unblock composefs and do not depend on the zstd:chunked timeline.