I have tried to QA the 500 GeV ZDC polarimetry runs.  If you go to the run page, for example, for 10075024 it is:

http://online.star.bnl.gov/RunLogRun9/Summary.php?run=10075024

and then go to pdf link you can see the plots.  I looked at the ZDC plots on pages 2 and 3, bunch crossing plot of page 4, and the BBC plots of pages 53-57.  I saw that the time differences for the BBC and ZDC were reasonably centered around zero - plus for the BBC that the vertex looked ok and east and west TAC reasonably agreed.  The list I started with is attached - Database.txt. 

A few things I found

1)I think all of day 70 should be discarded.  Look at page 2 of http://dean.star.bnl.gov/runPlots/10070004.pdf.  In all of day 70 the ZDC Time (West-East) plots were extremely off-center like for this run. 

2)A few runs below were not taken with a trigger that requires BBC+ZDC these include:

--10072112,10073029,10073032,70074004,10074005,10075011,10075012  These triggers did not require a hit in the BBC at all, but there was an additional requirement on the sum of the ZDC cal adc.  These should probably not be included with the standard trigger, but could they be useful?  What about the timing and a vertex constraint?

--10074008 and 10074009 were just BBC coincidence triggers.  This will not be useful for ZDC asymmetries (really low ADC hits in the calorimeters), but could be used to cross check the BBC analyzing power without a ZDC requirement, if the statistics are good enough.

3)A few runs had such low stats I should discard them.

4)For basically all of the runs the BBC vertex and east vs. west Tac on page 57 looked asymmetric/had some big scatter.  This occured for all runs, and there was nothing really really horrible so I think it's OK.

5)The runs that were taken like BBC East + ZDC East or BBC West+ZDC East were taken on day 67.  Given how bad day 70 looks, I don't think these are worth analyzing.  Plus, they are very low statistics.

Below is a plot of the filled bunches for fill 10375 for the data I analyzed.

the 3 bunches filled, 1 unfilled is consistent with cdev and the online plots.

The counts for east and west and the 4 spin states (++,--,+-,-+) are:

EW 0 Spin 0 Events 482934
EW 0 Spin 1 Events 477909
EW 0 Spin 2 Events 458314
EW 0 Spin 3 Events 446883
EW 1 Spin 0 Events 651129
EW 1 Spin 1 Events 640672
EW 1 Spin 2 Events 623567
EW 1 Spin 3 Events 615064

which seems fairly even to me.