Remove orientation_strategy from Modelchain

[kandersolar]

• Closesremove ModelChain.orientation_strategy #1028
• I am familiar with the contributing guidelines
• Tests added
• Updates entries to docs/sphinx/source/api.rst for API changes.
• Adds description and name entries in the appropriate "what's new" file in docs/sphinx/source/whatsnew for all changes. Includes link to the GitHub Issue with :issue:`num` or this Pull Request with :pull:`num`. Includes contributor name and/or GitHub username (link with :ghuser:`user`).
• New code is fully documented. Includes numpydoc compliant docstrings, examples, and comments where necessary.
• Pull request is nearly complete and ready for detailed review.
• Maintainer: Appropriate GitHub Labels and Milestone are assigned to the Pull Request and linked Issue.

Follow up to #1176 (comment). I went the lazy route and removed it without deprecation, but I am willing to not be lazy if a deprecation period seems like a good idea here.

[wholmgren]

Thanks @kanderso-nrel.

After some more thought, I'm leaning towards deprecating the parameter. The primary reason is that the default used to be south_at_latitude_tilt and users had to specify orientation_stragety=None to prevent ModelChain from modifying their systems (#290). As a long result of that bad API decision years ago, I still have modern code that sets the parameter to None. I'm guessing that at least a few other users might be in the same position.

FWIW, I briefly searched github and google for pvlib and south_at_latitude_tilt. I found one article in energies that used the parameter. I did not find any maintained code that uses it.

[wholmgren]

Forgot to add... On the other hand, the changes to the object layer make the deprecation non-trivial, so unless you know a good way to do it then it's fine with me if we skip it. We're not a 1.0+ project for this reason.

[cwhanse]

LGTM. I'm ok removing without deprecation, in this case.

[wholmgren]

Let's go ahead then. Thanks @kanderso-nrel