Update Chicago Health Atlas coverage files to have .json extension instead of .csv

[vingkan]

I made a mistake when I wrote the extract steps for the population dataset by naming the population coverage file .csv even though the contents of the file are actually JSON. We even read the coverages file as JSON in the transform steps, we just never fixed the file extensions.

Unfortunately that mistake has been copied into most of our other dataset extraction steps.

This issue is non-urgent, but we should rename those and update the documentation.

Don't forget to update all the places a coverage file for a specific dataset is mentioned:

• Extract step target name
• Extract step output file
• Transform step input file
• Transform step target dependency
• Documentation