Thanks, David, Tom, Tom and Glen.
I've seen those videos. Dave's tear down is really good though my wife makes me wear headphones for any of his videos - she's a non-fan. I kind of get her point...
After following the circuits on the schematic, I think the 38V/67.5V feeds are probably OK because there doesn't seem to be anything particularly sensitive there. Two 100n 50V ceramics are worth looking at, one on the X-Axis Amp and one in the vertical output amp. Everything else feeds though resistors. Though the lowest is 330 so bears looking at a bit more.
It looks like Q912 is not blown, at least a resistance check across the collector and base sees a cap charging up - probably C909 though the 2.2K resistor on the base. Not sure how to determine if either TL594 is dead. I've got the datasheet but haven't looked yet. Will probably order a few just in case, they're cheap.
On the caps, I've found what look like decent replacements - reasonably low ESR, same or better tolerance and all are 105C. However, not a single lead pitch matches. I can make them fit by bending the leads in an L to make them fit the original pitch. I assume that there is no issue with that. So, not worried except for C900 - the big 2200 uF 80V on the mains input board. The only matches are snap-in caps with fat, stiff, stubby leads and the wrong terminal pitch. (10mm vs a measured 12mm on the one that's there now). I guess I can make a little adapter but it will be a bit ugly. If I was industrious, I'd 3D print a foot for it to keep it from moving around - though it will probably be hot glued instead.
Another question. There are at least 9 10 uF axial electrolytics in various places. I should replace them as well? I'm not wild about removing the attenuator/timebase board, though it's probably easier than it looks.
Once I get my ESR meter (due this Friday), I'll hopefully find a smoking gun.
Anyway, again, thanks for all the help. I really do appreciate all your efforts.