The efficiency gap fails to guarantee or even optimize compactness, yet many gerrymandered districts are "thin slices" or strange shapes. (one simple defition of compactness is the ratio of perimeter^2/area)

What the author fails to do is explain why the efficiency gap is better than district compactness. If they are used together, then in what way are the two measures weighted? If not, why should we pick efficiency over compactness?

It should seem obvious that these two methods would result in different districts, but which mwthod to use is more a matter of personal values than objectivity.