Use hpmoc method to calculate sky map overlap integral

Using the current methods to calculate the sky map overlap integral, we iterate over every pixel in one sky map and look for the the closest matching pixel in the second. However this method could miss a significant amount of probability in the second sky map is very "pointed" (i.e. has a small region of very high probability density) and give an incorrect result (very possible with Swift sky maps).

This could be fix by using the hpmoc method, which produces a sky map with the highest recursion of each of the two input sky maps. We could then sum over these pixels prior to normalizing to get the sky map overlap.