I think this is misleading or misstated. Haversine distance of points on a sphere is a monotonic function of euclidean distance, so one's nearest neighbor is th

I think this is misleading or misstated. Haversine distance of points on a sphere is a monotonic function of euclidean distance, so one's nearest neighbor is the same in both. You must be warning people not to just take the sum of square differences on the actual latitude and longitude values.