Enabling customization of /or expanding downloadable file manifests

[CatLibWilk]

Currently the portal supports downloading a file manifest with an array of subject/object_id objects. It would be useful to be able to expand this manifest, in the same way that the contents of the 'data' download is determined by what is in the ElasticSearch index.

For instance, if files themselves are not stored in data commons, it would be useful to be able to select cases/subjects and then download a manifest of file_paths or file_names where data can be accessed externally.