Video upload metadata - document for future

# Summary
Currently Pipe stores some helpful metadata on videos. Obviously, when we move away from this service, we're going to lose this data. We need to get this data from the browser and store it in our database.

# Implementation
There might some questions that need to be talked out, but here's the general idea. We need to capture a few data points at the time of the video recording. The video metadata should only be present if the file was successfully saved to S3. Getting the data on the client should be best effort. The assumption will have to be that we can't always rely on the data gathered.

Data to capture and store:
- Client's IP address
- Browser and version
- OS
- Whether or not hardware acceleration is turned on (if it's possible to access this info) - see https://github.com/lookit/ember-lookit-frameplayer/issues/228

Open question:
- Should have a new table that's just for video metadata?

Start by tracing the API's video-recording save flow and the point where successful S3 persistence is recorded. Review how browser metadata can be passed and where video records are stored, then resolve the table or schema question. Done means the listed metadata is captured best-effort and persisted only after a successful S3 save.

Written by the indexing model from the issue text.