I didn't read past page 3 but people are arguing about charging up large reservoir caps, I haven't seen these in a PA5. Linear supplies have larger reservoir caps but the largest I have seen in the PA5 is 470uF after the switching transformer/diode. The PWM is determined by load and controlled via feedback. Input of 120v gets rectified into relatively smallish caps, nothing that's going to even kill a 30A surge current of a 1N4004. The rectified mains gets PWM through mosfets, the on/off current through Mosfets also passes through the transformer causing the reactive nature of large/small spikes depending on the PWM, this charges up the last cap which was 470uF, some 1,200uF. The feedback regulates the PWM which keeps the output stable under changing loads.
I typically see Mosfets failing not diodes, after that it's the largish 470uF-1,200uF cap after the Mosfets, transformer and diode that gets abused with 100kHz current pulses charging it up. They typically swell up and even explode sometimes. The Mosfets fail from on/off cycles that don't have lots of protection diodes around them. It's more likely the Mosfet fails short, and that causes diodes and other parts to fail. Other common failures I have seen is the input filters which prevent the power supply to pollute the mains, typically the common mode chokes are undersized and they get hot and expand, I have found cracks in both the soldering to the board and in the winding of the choke itself.
The PA5 failure complaints I have heard are almost overwhelmingly noise in the left channel first then failure of said channel. To me this sounds more of a problem with chips than the power supply. I don't think Topping takes advantage of the error reporting function of the chips and they also do not use the hard reset pin, this means the chips only can do soft resets. I'd have to poke through the chipset datasheet again to remember all the stuff Topping doesn't use that the chips provide but what do you expect from a cheap amplifier..
Power supply problems are real from repetitive on/off cycles but these are short intervals typically from your mains "flickering" when you are about to lose power or from a poor receptacle connection making poor contact. The back emf from transformer kills the mosfets due to no protection diodes around the Mosfets.
If you fail to follow the proper on/off procedure with these I can see the loss of mains before turning the power switch off a problem, The PWM of Mosfets is still screaming along at 100kHz yet the rectified 120v input is not there and the caps bleed down throwing the whole system out of wack from it's programming.