If that is too radical, maybe we can open up a working group somewhere on these forums where we can create a framework intended to become this core math library.
That might help adoption (especially if it is just offering bare bones things like trig) and speed up its inclusion with Swift.