As highlighted in #851 (comment), in the past year or so there have been several bugs and deficiencies detected and fixed relating to the concept of cyclic domain axes, and though the fixing PRs have included testing to isolate said issues to improve the level of testing, it is probable we could do with better testing since issues have congregated for the concept.
So, we should review areas where cyclicity should be covered in the tests but is not at present, and add sufficient testing in each such case.