Before I had checked our efficiency compares to the corresponding efficiency from the h-jet analysis:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-07052018-run-9-embedding-corrected-reconstruction-efficiency

To get a better idea of they compare, I plotted them on top of each other.

The band on the h-jet efficiency is (approximately) its systematic uncertainty, and the errors on my points don't mean anything.  Ours undershoots the h-jet, especially at low 'pT'.  The only parameters we have to vary are the max 'delta-r' and the max / min 'qT' cuts applied to the match candidates.  I don't see any good reason to deviate from our default delta-r cut of the resolution parameter (0.3 here), so I left that fixed.  I then varied the max 'qT' cut.

And all 5 on top of each other:

So it looks like I was being a little too harsh with my default cut of 'qT = (0.15, 1.85)'.