Here are the new calculated Run 11 gains:

Here are the old gains currently in the database (Run 9, pp200):

For reference...
Number of bad towers in 2009 = 154
Number of bad towers in 2011 = 171

For comparison, here are the differences new-old, shown in eta-phi (top) and eta (averaged over phi, bottom):

For comparison, here are the ratios new/old, shown in eta-phi (top) and eta (averaged over phi, bottom):

And finally, here are the relative differences: (new-old)/old, shown in eta-phi (top) and eta (averaged over phi, bottom):

Clearly there is a variation with eta: the deviations increase towards large eta.  Furthermore, there is a difference between the east and west barrel. 

With the TOF I calculate (DeltaInvBeta/InvBeta)|elec as described on slide 13 here: drupal.star.bnl.gov/STAR/system/files/PID_STARmtg_UCDavis_Masa_v3_0.pdf  If I plot DeltaInvBeta/InvBeta vs p I get: 

The real electrons are in the band around 0.  I assume that the band around DeltaInvBeta/InvBeta ~ 0.15 at low momentum are pions. 

I can then project into small momentum bins.  Here I show 1.5 < p < 1.6 GeV/c (binx=16) and 3.0 < p < 3.1 GeV/c (binx=31):  

In the lowest p bin there are two peaks, the second corresponding to the band seen in the 2D plot above.  My cut of |DeltaInvBeta/InvBeta| < 0.04 cuts out that second peak, as well as the long tails (pedestal?) in all p bins.