Update data quality metric outputs

[grgmiller]

As part of the research paper for OGEI, I wanted to update some of the output data quality metrics and figured these changes would be useful to merge for our next release.

This PR makes the following changes:

• Combines diba_imputation_performance.csv and national_imputation_performance.csv into a single file wind_solar_profile_imputation_performance.csv which contains one column for the DIBA results and another column for the national results.
• Adds a new output cems_pollutant_measurement_quality.csv which summarizes what percentage of CO2 mass, SO2 mass, and NOx mass reported in CEMS was directly measured versus imputed. This metric is based on the reported "mass_measurement_code" reported for each observation in the CEMS data.
• Adds additional summary columns to input_data_source.csv and hourly_profile_method.csv for CO2e, SO2, and NOx (both totals and _for_electricity values)
• Breaks out the EIA column in the input_data_source.csv to distinguish between monthly-reported EIA data and annually reported EIA data.
• Re-orders certain rows in the outputs so that they are generally arranged with "better" sources/methods at the top and "worse" ones at the bottom.

NOTE: I am requesting this to be merged into a new development branch where we can stage cumulative updates before releasing a new version on main

[gailin-p]

Can you confirm that this works with data_pipeline.py --small True? Some previous per-method data quality metrics failed because the --small run doesn't use all of the methods, so just want to confirm that this avoids those issues

[grgmiller]

I have confirmed that this works with --small in 2020