That did the trick - well, mostly. The tab by the volume control just would not bend enough to release & ended up breaking. But the top is off, the receiver PWA is out, and the scene of the crime is accessible!
R4 is marked '100', i.e. 10Ω, but is clearly burnt and measures 54.5Ω. The HEF4013B measures about 1.5Ω Vdd (+ve) to Vss. I will replace the IC, and change R4 for a series R-LED combination to drop the supply voltage a bit, there's plenty of room. I need to trace out the circuit that drives the power switching first, to check for issues.
The 2SD882 power transistor on the amplifier PWA doesn't appear to be the switch, it looks like a capacity multiplier (to reduce supply ripple) but quite a crude one (no base stopper?). If so, a suitable resistor from base to ground would reduce the peak voltage to the audio amplifier.
I might possibly stick a resistor across the 18V smoothing capacitor to help limit the peak volts a bit, but I can experiment with that once it's all back together, using the Variac (actually mine is a Regavolt). 'Power Off' draw from the supply I measured at 0.87W, so we could sacrifice perhaps another 0.5W in a good cause, if it's effective.
Enough for tonight though, it's been a long, hard day!