The page makes a good case that the method is more regular than other approaches, but then doesn't compare its performance to naive regular approaches like latt

The page makes a good case that the method is more regular than other approaches, but then doesn't compare its performance to naive regular approaches like lattices or distributions derived from standard aperiodic tilings. > need to distribute points on the surface of a 3-sphere as evenly as possible The examples there are of a 2-sphere.