I realized yesterday that the profile on the response matrix I had shown to JetCorr was using the wrong axis:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/jetcorr-follow-05292018-tprofile-check-response-matrix

What I should be doing is calculating the spread for each bin in x.  What I was doing is calculating the spread for each bin in y, and then plotting it along the x-axis.  Below I show the difference in the x and y profiles using the response matrix obtained from our new jet matching algorithm.

And with a different binning: