If you set the body length, arch height, fingerboard projection, saddle height, and bridge position and height, then the total string angle is whatever it is. I do those other things, and don't do anything too abnormal, so the total string angle comes out about the same (normal) every time. (edit: oops, I neglected overstand, which also affects string angle). If I did setups on short body / high arching violins or other oddities, I would probably look at total string angle.
I don't know of any theory that can determine an "ideal" string angle, other than what seems to work best in practice.