As someone who is firmly on the math side and knows very little about music theory, it seems like you should never expect music theory to answer subjective questions like "Why does x sound better than y?" There are so many factors for why someone will find a sound pleasing or not, and most of them are biological/psychological.

IMO asking why a melody sounds pleasing would be like asking why a plot of the Mandelbrot set is interesting; math isn't going to be able to answer that.

I would be fine with a music theory that is a set of tools to deal with sounds, just as math is a set of tools to deal with numbers. Whether certain properties are desirable are up to the theorist.