Maybe not addressing all six frame buffers correctly.
I looked around similar offsets to the range of buffers we saw in the ipl getimepath 0 results, and I found 16 buffers with live data, a few before the starting buffer shown in the 656x480 P2 pipeline, and several after. I tried filling all of them but still have a flicker. I'm not going to dwell on that issue though---pretty neat to have a histogram regardless, thanks to
@0dan0 !