What level of error do I encounter when approximating the earth as a sphere? Specifically, when dealing with the location of points and, for example, the great circle distances between them.
Are there any studies on the average and worst case error compared to an ellipsoid? I'm wondering how...