Other than from defective tubes, it is not normal for cathode resistors to simply blow out, let alone on apparently multiple examples of a given unit. Therefore, there are likely design issues that are encouraging arc events to occur. With that thought in mind then, here are a couple of observations to possibly help solve the problem for all users of these amps:
1. The 180 ohm screen resistors appear to still be in tact, or if changed, it does not appear that the originals ever burned out requiring replacement, indicating that the source of the damaging cathode current is most likely from the plate in this (rather rare) instance. The screen resistors are hardly placed at an optimum position (very close to the screen terminal), but in any event appear to be doing their job.
2. Long OPT primary leads and traces on the circuit boards are always problematic relative to output stage stability -- particularly when a push-pull-parallel design is involved. Therefore, I would recommend a little circuit board surgery to help prevent such occurrences in the future:
As close to the plate pins as possible (pin #3, and preferably right at the pin), break the trace leading to this pin. Then, solder a 22 ohm 2 watt resistor across the broken trace -- that is, from pin #3 to the side of the trace leading to the OPT. This effectively places the resistor between the plate connection of the tube, and the plate connection of the OPT right at the tube, and will need to be done at each of the 8 output tubes in the unit.
Finally, the cathode resistors are correctly sized at a 1 watt rating, since the dissipation through these resistors is just under .5 watt at full sustained power output. With proper output stage stability then, unless a catastrophic event in a tube takes them out, 1 watt cathode resistors should last for the life of the unit. Installing the individual plate resistors will go a long way to ensuring that stability exists.
I hope this helps!
Dave