I am about to blow a fuse here trying to interpret the difference of signal between Ch1 and Ch2... The right channel looks lower output than the left, but....
When I see the signal on the screen, it looks like Ch2 is smaller... but if I invert Ch2 and then add the two signals together, then boost the sensitivity of the reading and/or increase the signal from the function generator.I figured it out, but now looking at the difference of signals, the left channel being normal, starting on a rise and crossing the 0 at the zero:

As I turn up the output, I get what I initially expected, the left channel slightly pulls the signal up from zero (at the left side of the axis)...

as I keep increasing volume, however, suddenly the inverted right channel signal becomes dominant and the inverted signal becomes dominant...

What's *really* weird, is that this happens not only when I do the output adjustment, but even when I increase the volts/div sensitivity of the scope.
If someone has the patience to interpret this, and either explain something in my knowledge that's missing, or has an explanation for this phenomena, I would be very grateful. I'll post pictures/a video to go with this post shortly.
[edit update]
I flipped the board over and noticed a number of bothersome things: when I did the recap I didn't do a particularly good job. I only had a flux pen and not actual flux I could apply onto the board. As a result there were some extra solder whiskers floating around and a number of my solder joints could probably have been better.
I went in and cleaned all of them up, reflowed almost all of them, re-checked the biases and repeated my measurements from before. I'm not sure it electrically made a difference, because I didn't take notes of my measurements, and the phenomena still arises where the left channel pulls stronger at lower voltages (~2v) and as I crank it up to ~20v output the right channel pulls stronger (when taking the difference between the channels).
However.. measuring with the Fluke 45, I get 20.12v on the left and 19.9v on the right (this is maxing my function generator when attenuated -20dB).
This doesn't make sense. When looking at the difference of channels, and inverting the right channel, as I crank the voltage up, the waveform begins to follow the inverted right channel suggesting it is "stronger". My ears agree: the soundstage sounded shifted to the right. The Fluke 45 DMM disagrees. This phenomena stays even at different frequencies, although my ears hear a frequency dependence to this phenomena.