[Spike]: Make major design decisions for spreadsheet regeneration

[justincc]

See Ingest Additions and Analysis AUDR HLD 2019Q3.

These are major decisions such as:

a) Do we use data from the datastore or ingest to populate these regenerated spreadsheets?
b) Is it okay to generate a spreadsheet which contains all the data for a project or will this be too much? Does this need the user to be able to select what to populate?

Arathi has been working on spreadsheet regeneration so they can test new schema changes immediately on end to end test. Need to coordinate with them.