[Argus25]

It will be interesting to ultimately find out why the oscillator is failing, since most of the obvious suspects have been eliminated and we know that it once worked properly.

Hopefully it will be "one defect" responsible.

One horrible outcome that can occur in a case like this, is a collective accumulation of small losses that conspire to give the defect, like dielectric losses, core losses , low gain or reduced transconductance in the valve, slightly low heater voltage etc etc. It's like the remark where someone says: "That's a lie" and the response is "No, its just an incremental accumulation of half truths"

Then in that case its hard to know what to target for a definitive repair, so I hope it turns out to be "one thing"