I think that 1.32 derives from the assumption that the drag curve is the sum of the profile drag curve, a pure x squared curve and the induced drag curve, a pure 1 over x squared curve. In that case the tangent from the origin touches at 1.316 times the x value at min drag. (4th root of 3, mathematically) Of course, in real life these curves are not pure curves, so 1.32 is a rough approximation, and goes seriously wrong when, eg. wave drag upsets the drag curve at higher M.
Dick W