Sorry for causing the regressions and thanks for the fix!
I just checked all the no-meta files. The reason that they did not cause the tests to fail is that they still have at least one tag defined that is not read (encoder settings for example). I think it would be better to completely remove the tags from the no-meta files instead of adding another test file.
You can easily do that with the kid3 tag editor, but I can also do that if you prefer.

In D13885#287042, @astippich wrote:

I just checked all the no-meta files. The reason that they did not cause the tests to fail is that they still have at least one tag defined that is not read (encoder settings for example).

Okay, so can confirm that what I wrote in the description/summary is correct ;)

I think it would be better to completely remove the tags from the no-meta files instead of adding another test file.

Fine with me. I did not spent time thinking about whether the almost tag-empty files are covering proper test cases or if they should have been really empty, as in tag-free :)

You can easily do that with the kid3 tag editor, but I can also do that if you prefer.

I used id3v2 -D test.mp3 (fixed: -D, not -f) to create the test.mp3 without any id tags from the existing :) But had to goggle up how to do that, so happy to leave this to people who have experience :)
So happy to have you take over this patch, all I want is to have the tests fixed :)

All updated no-meta samplefiles fail for me without the patch, but as wanted do not fail with the patch. So seems fine to me :) @astippich Thanks for picking up this patch and creating proper test files.
+1 though only, given I am no maintainer.