Debug issues with PudlTabl output objects and direct to SQLite outputs

[zaneselvans]

The database produced by the direct-to-sqlite ETL is (at leaset) slightly different from the old datapackage based database, and this is causing issues with the pudl_out / PudlTabl class that we rely on to pull denormalized data tables from the database. We can't run the tests without the outputs working, so this means we don't know how broken everything is with the new DB 😅 See #1195

Known problems

• [x] Many of the output objects complain because there's no id field in the new database tables (this includes at least pudl_out.plants_eia860, pudl_out.gens_eia860, pudl_out.frc_eia923, and pudl_out.bf_eia923)

[zaneselvans]

Thankfully, once the tests were up and running, it seems like all the current output routines, mcoe etc. work fine -- just those ID column references needed to be removed.