My motivation here is that this API pairs well widthunicode-width (or with any terminal that uses similar rules tounicode-width for determining character advance width, IOW ~all of them). Defective combining character sequences are one of the main cases that these width APIs don't handle properly, so correcting them beforehand is useful. (#97 has a similar motivation.)

In that case please include some description of this API's use cases in its documentation.

I still don't know if or when I'll be able to review this and the other PRs.