More consistency in ETL functions regarding data["attribute"] vs data.get("attribute")

[mbertrand]

Description/Context

Some ETL transform functions use the format source_data["attribute"] for some/all attributes- this is likely because the attribute is always supposed to be present in the source data, and the function should fail if it isn't, to indicate that something is wrong with the source data.

Other functions use the format source_data.get("attribute") for some/all attributes - likely because the attribute might not always be present in the source data, and this is considered okay for the particular source.

It might be good to double-check on these and make sure they are still used appropriately per data source and per attribute, and make them more consistent where applicable.