Efficient computation of arbitrary beam scattering on a sphere: Comments and rebuttal, with a review on the angular spectrum decomposition