I've been looking at the sizes for the TPC data volume spikes.  

I checked out 3 runs:   20041001, 20042002, 20042013.     During the 10am mtg, the thought was that this only occurs on the east TPC.  This is not true, the issue shows the same type of signature (with different patterns) on both the east and the west.    However, each time, the full pattern appears on the east or on the west.  Not on both at the same time.

The inner and out sectors are both affected at the same time

Here I plot the sector vs the the time for each run.   I only plot the events greater than various thresholds (and the higher thresholds are plotted later, so the dominant symbol is the highest threshold satisfied.)   I offset the outer sector by .5, so adjoining inner sector and outer sector appear next to each other.

The bands where sectors are sending large amounts of data are of course perfectly correlated with Tonko's data rate plots, with the exception of the first short band in run 20142013, which is too short to make a contribution because of averaging.

In run 20041001, the issue only affects the east side, although you can perhaps see some effect on the west in the outer sectors which have higher gain.   In both of the other runs the west and east are each affected, but at different times.

The magnitude of the effect can be categorized into 3 groups:

WORST  (>40k pixels hit):     Sectors 5,6,7         and   Sectors 14,15
BAD:     (>20k pixels hit):     Sector 9                         Sectors 13, 16
Lowest   (>5k pixels hit):      Sector 4, 8

None of this seems to map in any meaningful way to the Gating Grid drivers...