Those reflected beams would at best be very very weakly entangled, right? I'm not sure what's the name for it, but if you arranged the quantum state of the two beams into a 2x2 matrix then the determinant would be just a tiny bit non-zero.