Ok, the transformer output is in the form of a sine wave, actually two sine waves that are 180 degrees out of phase. Each red wire from the PT secondary carries one of them, and the red/yellow wire (the center tap) is common to both. The rectifier tube contains two diodes, which only allow current to flow in one direction. The diodes essentially eliminate the bottom (negative voltage) half of each wave, and then superimpose them at pin 7. What you would have at pin 7 without the 20uF capacitor is a waveform that would look like a series of humps, from zero to some peak positive voltage. The capacitor functions like an energy storage reservoir, the same way a flywheel turns the discrete power pulses of a reciprocating engine into relatively smooth shaft torque. If the reservoir cap is incapable of storing sufficient charge, there will be a lot of the pulsating character (ripple) left. I suspect that the "red" section of the multi-cap is nearly shorting out the new Xicon.
The first thing I'd do is cut the red lead from the multi-cap and see what that does.
Electrolytic caps have a finite life-span, and it's a pretty safe bet that yours are past due, and the other two sections of the multi-cap are probably not within spec. If you want to preserve/restore the original character of the amp, I wouldn't change the values from those shown on the schematic too much, not more than 10-20% say, as the amount of filtering will affect the tone.
To replace the multi-cap completely, you can leave that 22uF Xicon where it is, and find some way to wire and secure two 10uF caps somewhere else. You could also make up a bundle of three new caps, and find a way to use the clamp or just the mounting hole that was used to secure the original to tie the bundle down. Or, you might be able to find something close to the original such as here:
http://www.vibroworld.com/parts/tech15.html#cans
There is a 1" diameter 20-10-10 can listed.
There is a 10-10 with wire leads here:
http://www.tubesandmore.com/scripts/fox ... I%26SINGLE
This also applies to the cathode bypass caps like the 10uF on V4. That cap boosts the gain of the oscillator, and it may have degraded enough that the trem won't break into oscillation any longer.
I'd replace the 47uF at V1a's cathode with the correct value just to restore it to the original schematic.
