
Meanwhile I measured the voltages on the filter caps : 291V--285V--249V (vs. nominal 293-287-252) which is within 1% so that looks OK. Checking the circuit diagram again I see there's a couple of other voltages I could have measured but most likely they're OK too. I also tightened up the springs of the tube sockets and swapped in the new output glassware, GE 6005/6AQ5W's, and things are already sounding better.
The electrolytics are somewhere in the christmas post, supposedly dispatched on 18 and 19 december, with a bit of luck they should arrive while it's still 2018. While waiting I found this video : which is interesting because the amp does fizzle a bit and my el cheapo DMM does show "something" when set tp its 2000K range - "something" because the figures change faster than I can read them. So maybe I'll change this one too, and the question is of course - stick with 10 nF or go up to 22 or 47?