Recently, I tried unfolding the pi0 data using only the embedding sample.  The details can be found in these posts:

https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-08162019-run-9-pp-new-r-02-pi0-unfoldings
https://drupal.star.bnl.gov/STAR/blog/dmawxc/update-08192019-run-9-pp-pi0-unfolding-systematics

Here, I try to unfold the gamma-rich data also using only the embedding sample.  I tried out two different background-subtraction schemes:

1. Fully unfold gamma-rich data, and then subtract the fully unfolded pi0 spectrum from the fully unfolded gamma-rich spectrum
2. Subtract the raw pi0 spectrum from the raw gamma-rich spectrum, and then unfold the gamma-direct spectrum

They yield consistent results, but -- as one might expect -- the uncertainties are substantially larger in method 1.  The list of systematics is the same as in the pi0 unfolding.

===========================Method 1 ========================================

eTtrg = 9 - 11 GeV:

eTtrg = 11 - 15 GeV:

eTtrg = 15 - 20 GeV:

I'm not sure what's going on in the 15 - 20 GeV bin... :/

===========================Method 2 ========================================

eTtrg = 9 - 11 GeV:

eTtrg = 11 - 15 GeV:

eTtrg = 15 - 20 GeV:

It looks like something went wrong with the 11 - 15 GeV subtraction... :/