This is screaming "clocking issue" but without a scope, it will be very difficult to determine what went wrong. The first thing that I would suggest is to NOT use the config utility to check for communications. This forces the chip to write to it's internal memory and that can corrupt the chips config and every time you reflash it, you take more life out of the chip.
In order to get an idea of what the issue is, I would remove ALL of the PIC's from the malfunctioning Zeus. Then I would get a configured, functioning, and tested PIC put it on the non functioning Zeus by itself and lets try it in socket 4. Then connect up the string of nodes that you tested the PIC with to that port, connect up the ETD and power supply, remove the program jumper and see if the PIC functions using sequencing software such as Vixen or Xlights running a test patterns and NOT the config utility. If not, move the PIC and string to a few more positions and see if any of them work. If not, then there is a pretty good bet that either the clock oscillator has failed or the signal is either being blocked or interfered with. If it does work, then start adding PIC's one at a time and retest after adding each one. If it stops working after adding a PIC, try swapping that PIC with another one and see what happens. If it works with the different PIC, that PIC is probably defective and is killing the clock signal and if not, then either there is something wrong with that position on the board or there is something pulling the signal down to a point where the load of X number of PIC's pulls the signal down too low for the PIC's to clock properly.