Update pvgis test data files

[kandersolar]

• Closes #xxxx
• I am familiar with the contributing guidelines
• Tests added
• Updates entries in docs/sphinx/source/reference 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 (including remote-data) and Milestone are assigned to the Pull Request and linked Issue.

PVGIS TMY data has apparently been updated again, causing remote-data tests to fail. https://github.com/pvlib/pvlib-python/actions/runs/14386680023/job/40343717051

This is the latest in a sequence of changes needed to keep the PVGIS tests passing. Perhaps we should rethink how we test these functions. Previous PRs:

• Update PVGIS TMY tests and files #2222
• Fix PVGIS TMY tests #2305
• Fix get_pvgis_tmy for outputformat='csv' #2395

[AdamRJensen]

I agree that it's not optimal. Though at the same time, some of the changes to PVGIS in the past has required that we modify our functions, which we wouldn't have discovered if they didn't fail.

It's a shame that PVGIS isn't better about consistency though.