I don't think there is a problem is the LVPS.
I don't think either. Perhaps I wasn't clear enough: I think the LVPS is going into protection mode after the "snap" event but the cause is somewhere. Perhaps the LVPS going in protection mode causes the "snap" you hear. Sometime capacitors especially tantalums fail intermittently and unpredictably and becoming basically a quite low resistor, like 50Ohm, which trips overcurrent protection in the power supply making it hard to detect because they don't continue cooking off and releasing smoke.
One of the output transistors gets pretty hot while the other does not.
there you go... the one which is not getting hot is the suspicious one, but check all four Q13-16, the resistors and caps as well once the transistors are out of circuit.
I am not very good at visualizing circuits so I don't know how the signal path works
reverse engineering is a crucial skill but before you get efficient at it you must learn forward engineering and the basic building blocks. It'll come with time and it's a muscle you have to excercise. I suggest you watch as many "back to the basics" videos (also the other are worth) as possible from w2aew and Mateo Aboy if you want to learn the basic building blocks of circuits and how they couple together. Taking a proper course would also help
In the meantime, just ask... we can go through a section you don't understand. I can't promise accurate analysis: I'm just a hobbist but I have a degree in engineering and I learned some theory some years ago.
it is very hard to get a scope probe onto most of A13 to trace things.
That can be hard. Take your time to set up probing eg. solder a lead wire to make a signal accessible. Probing and measuring is the only way to understand what's going on so don't dismiss it as a quick and easy thing to do. Also record the results, maybe take screenshots of the measurements, and take notes / record yourself (may seem silly but I do it all the time).