The first simple step would be to compare the overlaps of the five notable maps with each other vs. the baseline overlaps -- first for NC and then make a few other states. (or all, if it's a simple script)
A more rigorous step would be to get an ensemble of randomly generated maps for a state (e.g., NC) and do the same analysis for the six baseline variants (see #63 and #64) plus the maps in the ensemble.
Do this empirically.