This has been a right nightmare to diagnose and fix but finally got there.
Ended up de-soldering and putting in sockets for 20 IC in the audio Circuit.
Was it any of the 4 amps? ...of course not (well more on this later)
Was it the Sound CPU? ...Nope
Was it the Sound Ram? ...Nope
Was it the Sound ROM? ...Nope
Was it either of the two Yamaha Sound IC? Yeah, yet another dead end.
Was it any of the TTL 74LSxxx Logic chips in the sound circuit?
Nope as all of them test good out of circuit apart from that puzzling behavior of the mirror outputs of the two 74LS374.
I had to remove the socket i had put in to figure out what was going on and found that there was a broken trace between Pin 19 and Pin18 of the pair which I repaired. But the Pin 15 and Pin 14 connection was weird if connected. Well the simple matter was that its not connected at all on this bootlegs PCB by default, unlike on an original board. So I left it disconnected.
And this leads me to eventually going round all the board looking for broken traces in the sound section and finding nothing irregular.
Ages ago I had tested IC5, a 74LS240 that feeds the Sound CPU its 3mhz clock. I had only probed with my basic oscilloscope to ensure it had the correct frequency on its input and output pins which I did and thus I ruled it out.
However while scouring the schematic I noticed that Pin 3 of that 74LS240 is actually also connected to Pin 1 of a 74LS155 in the sound logic and low and behold its output looks strange... How did I miss this???
So assuming a bad IC I remove it to find it only blooming tests good as well...
But wait... I find a corroded trace underneath the IC..
Which I repaired and low and behold the sound comes back!
Albeit with very bad hissing in the Background Music so I replace the YM3014 at IC23 and its all sorted.