There's no issue with the WireGuard® implementation aside from the fact that the author of WireGuard®, Jason Donenfeld, didn't like that it wasn't a straight port of his code, but was instead a new implementation based on the WireGuard® protocol paper.

He never articulated any reasons about what might be wrong with it aside from vague handwaving. The implementation actually works well.