Carl Worth provided a nicely detailed analysis of the issue on the upstream bug. The gist is that as per the spec, this behavior when using zero-width lines is permitted to differ from driver to driver, and it is up to wine to use width=1 lines instead of zero-width lines.
I will forward this bug to the wine package for further handling, since it is not considered an X driver bug.
Carl Worth provided a nicely detailed analysis of the issue on the upstream bug. The gist is that as per the spec, this behavior when using zero-width lines is permitted to differ from driver to driver, and it is up to wine to use width=1 lines instead of zero-width lines.
I will forward this bug to the wine package for further handling, since it is not considered an X driver bug.