We don't need a separate specialization for ints (we can rely on the compiler to inline isnan as always false), and having a separate loop to find the first non-nan value isn't necessary either.

One change in behavior this causes (in this private function) is that a fully nan input is considered as being sorted, but this is consistent with empty inputs being considered sorted too, and also OK with the actual use of the function (determine whether to go through a fast path in drawing lines -- fully nan lines are not drawn anyways).

followup to #24575, which I unfortunately have to slightly revert (going back to x != x) due to the lack of isnan(IntegralValue) on MSVC (https://stackoverflow.com/questions/61646166/how-to-resolve-fpclassify-ambiguous-call-to-overloaded-function), even though cppreference declares that it should exist (https://en.cppreference.com/w/cpp/numeric/math/isnan)...

PR Summary

PR Checklist

Documentation and Tests

• Has pytest style unit tests (and pytest passes)
• Documentation is sphinx and numpydoc compliant (the docs should build without error).
• New plotting related features are documented with examples.

Release Notes

• New features are marked with a .. versionadded:: directive in the docstring and documented in doc/users/next_whats_new/
• API changes are marked with a .. versionchanged:: directive in the docstring and documented in doc/api/next_api_changes/
• Release notes conform with instructions in next_whats_new/README.rst or next_api_changes/README.rst