Keep detailed Content-ID fingerprints as ISCC metadata.

[titusz]

The algorithms for the Content-IDs make use of more detailed media-type specific fingerprints. Currently these fingerprints are discarded as soon as the Content-ID is calculated. For use cases where a more detailed and robust matching of content is a requirement it would be valuable to keep those detailed fingerprints. We could extend the API so that the code creation methods return those detailed fingerprints. We could also specify a schema for encoding and storage of detailed fingerprints as identifier metadata.