They still don't know what's driving the 1/2-wave plate correlated widths.
From Elog 1467565 (R_Suleiman):
Scott Higgins just mentioned to me that the Helicity Delay was set to 8 in the Fast FB code but actually it is 0. This number is hard-coded and has to be updated each time we change the Helicity Delay. That makes sense; the measured charge asymmetry was centered at zero although Hall A was not doing charge feedback. Scott just went ahead and change the code to use 0 Helicity Delay. Still, we have to find out why the fluctuations change between IN and OUT.