Any idea why there is such a discrete separation between the near-polar orbits and the 53-degree orbits? Naively I would have thought that there would be sever

Any idea why there is such a discrete separation between the near-polar orbits and the 53-degree orbits? Naively I would have thought that there would be several classes orbits with degrees ranging from ~0 to ~90, with the number in each chosen so that coverage roughly matched the subscriber distribution.