[studies] Weird working of the config file for studies

[jgbarah]

Apparently, when specifying no studies (eg, studies = []), the corresponding section for specifying the study arguments is still needed. My impression is that only sections for active studies should be in the config file. That is, both the following configurations should be valid:

studies = []

studies = [enrich_demography, enrich_areas_of_code]

[enrich_demography]

[enrich_areas_of_code]

Right now, apparently only the second one is valid.

[jgbarah]

Closed via #157