Very slow check-in of master table

[seasidesparrow]

The data check-in process is very inefficient, and is resulting in a bottleneck for setups where the database isn't running on the same server as the container.

• Grab all comparison data from the server at once, instead of row-by-row
• Write all valid checkins and failures to lists
• Commit both data blocks to db at the end of the process, not row-by-row