Design document on where to store data backups / data to be imported

Most data is available, but we probably don't want to be hitting these external sites too often. At the same point, hosting large files ourselves costs money. Also, it could be the case that we process a lot of raw data and store the results in the DB - what do we do with the raw data, if anything?

Clear guidelines and best practices are needed here. Doesn't need to be perfect the first time around, this isn't the final version!

No files, tests, or entry points are named. Start by reviewing how the project currently obtains external data, imports it, stores processed results in the relational database, and handles raw files. Done means a first-pass design document with clear guidance for backups, external-source access, hosted files, and raw-data retention.

Written by the indexing model from the issue text.