Consider file reading REST API

[zachmullen]

Current file reading API:

• GET /files/{id} to retrieve metadata
• GET /files/{id}/download to retrieve the blob

We should consider an alternative that is closer to S3/minio semantics:

• HEAD /files/{id} to retrieve metadata
• GET /files/{id} to retrieve the blob

[zachmullen]

FWIW I'm not necessarily advocating for the second one here. Just want opinions.