...
This hypothesis is further strengthened and proved if we see the results in the attached sheet Analysis Results. It's clear from the result that for all those targets for which Error Distance is less than 1000Km, the MinRTT of the first selected landmark is less than even 60ms. This means that if for any target we are able to find a nearest landmark (with MinRTT<60), then we will be getting good results and smaller this minRTT is better are the results. From the attached results, we can also see that for all those targets for which we got error distance above 2000Km, the MinRTT (column D) was above 100ms.
From these results, we can roughly conclude that if we mange to have our landmarks distributed in the world in such a way that we get low MinRTT for any target, the error distance can drastically reduce.