WIP: re-enable mypy for pySHiELD

[romanc]

Description

⚠️ Work in progress ⚠️

This PR re-enables mypy (not only for the source code directory) in pyshield. It then fixes the obvious issues and leaves two things to discuss:

1. In PBLConfig, should dt_atmos be of type int (as type-hinted now) or is it supposed to be a float (like other usage of dt_atmos would suggest?
2. translate_pbl_subtests is claiming that pyshield.constants contains a constant called CDTN, where such a constant doesn't exist. Should that one be imported from pyshield.stencils.pbl.constants?

@oelbert I guess you are the expert here. What do you think about the question above?

This PR is stacked on top of #81 and should be merged afterwards. This also explains the pyshield-side of failures in NOAA-GFDL/pace#173 and why we haven't seen them before.

How Has This Been Tested?

All good once CI is green.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
  See inline comments
• I have made corresponding changes to the documentation: N/A
• My changes generate no new warnings
• Any dependent changes have been merged and published in downstream modules: N/A
• New check tests, if applicable, are included
• Targeted model, if this change was triggered by a model need/shortcoming: N/A

[romanc]

This is why mypy never complained because that's a case-sensitive file filter that wasn't updated when the source directory was renamed from pySHiELD to pyshield.