Jeffery - I would encourage you not to throw in the towel. Your experience is impressive and there is no doubt that you know what you are doing. My comments are intended to perhaps stimulate your thought process in seeking the cause of the problem. To that end, it would be helpful to be able to replicate the problem without damaging the amp further. If the fault results in damage, look for ways to prevent it. You are already using a DBT to limit line current, and you can also combine that with a Variac to reduce the line voltage, for example.
If there is no combination of input power limiting that both replicates the problem and allows the amp to come up without damage, consider changing fuses to lower amperage parts, just for testing. Naturally, such fuses may blow during startup due to the inrush current for power supply caps. Then you have to get trickier, paralleling lower amp fuses temporarily and clipping some out hot once the amp starts up (very undesirable). Just an idea.
There are undoubtedly other and likely better techniques, but the point of course is to repeatedly reproduce the problem while monitoring points of interest, without causing damage. IMHO, your oscilloscope is the tool for capturing transient faults. The scope will patiently wait for a drop-out or other event and if the amp can survive the aftermath, you can methodically trace the fault without babysitting the process each time.
As difficult as this unit has been to repair, it will be that much more satisfying when it is repaired and playing.
