It seems the ps h/s is heated by the internal air as it is near the fan,
rather than cooled, but presumably within allowable parameters, cannot see
any retrofit way around that.

So with the top removed and replaced with a large book, a thermometer over
the PA hs and another over the ps hs and one in the fan exhaust, ambient 20
deg C

All controls at mid and 4R dummy load
Initially 12.04V ac out for 6 minutes and then reduced to .1V input/ 400 Hz,
and 7.47V ac out,
reducing input to 1/10 while removing top and reading the thermometers for
that 30 seconds or so and loss of internal heat
6 minutes , 11.93 Vac, pa 72 deg C, ps 33 , exhaust not yet monitored
input reduced
15 m (6+9), 7.37V, 72, 45, 33 deg C
23m, 7.34, 72, 47, 36
30m , 7.33, 69,50,36
40m,7.32,77,52,37
50m,7.32,76,53,37
55m , 7.32,75,53,37

I reckon if the ps hs was not being heated by the internal air it would have
been only 35 to 40 deg C

No sound distortion emerging . So will just make those socket changes and
few paliatives like removing the paint on the chassis around the pot bushes
and resolder the usual suspects