Improve SQL queries for index and shelve scripts

[klassenjs]

INFO- Fetching region lookup from Danco - takes about 5 minutes because it is (likely) reading the whole table (1,500,000 rows) into python memory for use later to join against generally a few thousand pairnames. We can do much better.

[clairecporter]

Yes, we can use the microservice that Danny set up to get the region. Although it will work better when Rory makes the changes he's working on to allow multiple items per API call. Other approaches welcome too.