I think the easiest approach would be:OK, that helps a lot on the supply side. The diagram shows a common supply for the two halves of the dual PL8, but it still doesn't address the issue of 8 cell segments which are hard connected in series. I think I have a safe solution for you, but I would like to see Ian weigh in on this also. He is considerably more experienced than I am.
Take the time hit and do 6s cell blocks per PL8 .. with the bus bars removed so that only 12s is connected.
Alternatively .. the one way to do the 8s approach (to be able to get through testing faster) .. and retain safe isolation .. might be something like ...
Start with the buss bar attached between two 12s modules .. into one 24s block.
Do a test of the 8s cells in series across the middle .. 4 on 12s-ModuleA and 4 on 12s-ModuleB .. careful where you connect +/- to.
PL8-A can do this on one 24s block , while PL8-B does the same on a different unconnected 24s block.
With the middle 8s done out of the 24s , that leaves 16 cells left , in two 8s.
Remove bus bar between modules .. to make 2 separate 12s blocks.
PL8-A test side can do 8s cells on ModuleA
PL8-B test side can do 8s cells on ModuleB