Add ability to name a different default config file for subset_data

[ekluzek]

I need the ability to name and use a different default config file for subset_data. This allows you to use subset data for example to use 1850 datasets, or a different resolution file.

I think eventually we plan to get the surface datasets from a fake_case that subset_data creates and that would supersede this usage. But, until we have that in place, this provides more flexibility that's easy to put into play.

This is required for #1674