The inconsistent use of VOLUME_NAME_NT and VOLUME_NAME_DOS was addressed incidentally in issue 29734, but that issue is primarily about the handle leaked when GetFinalPathNameByHandleW fails. That said, you've cleanly addressed the handle leak in your PR also. I prefer the common cleanup approach that you've used here, rather than spreading the cleanup tasks around in blocks, where it can easily be overlooked.

The PR for issue 29734 also fixes some handle leaks related to win32_xstat_impl. Those leaks still need to be addressed for 3.7, and backported to 3.6.