The clipping at the top and bottom of the sine wave is coming from the generator I think.  When the generator is loaded too much by the attenuator, it clips.
To simulate a pot at the generator output, when one resistor increases value, the other needs to decrease, and vice-versa.
Otherwise when the resistor going to ground gets too low value, it loads down the generator (if the resistor connected directly to gen. is not changed).

The other little notch in the waveform is not exactly crossover distortion as it is not in the center of the wave, at 0V crossing.
It may be the turn on point of the secondary output transistors (Q15 and Q16) which is something loudthud mentioned back in post #188 :

Quote from: Loudthud on November 26, 2024, 07:42:24 AMA crossover notch might not be exactly at the zero crossing but at a slightly higher Voltage where the second pair of output transistors turn on at around +/- 3V.

In any case, it is still affected by the bias adjustment.

Without looking at the schem.,I would guess that the two sides are the same, so you should be able to compare readings on transistors etc. to the good side.

Like Miyagi stated, idle current is with nothing plugged in to any input.  So you do not need to check voltage across those resistors any time a signal is applied.

Your load box is connecting the amp output jack to chassis ground, via the scope BNC connecter.  That is bad as it shorts out R63, which is part of the feedback loop and will affect the sound.  It may also cause instability which could be messing with the idle current.

Use your scope probe.  Connect scope ground to chassis, and the scope probe to the red binding post on your load box.
When bringing up the voltage with variac, you may get idle current surges like that acros the 0R62's until it reaches a stable voltage.  Are you sure it was not doing that before?

Do not proceed with signal if you have a problem with idle current (no signal).
Get that sorted out first.
Disconnect anything plugged in to input or FX return.
It is critical that the (-) side of the output jack does not get connected to chassis ground (because R63).  Sometimes the test equipment can do this.
Make sure there is no DC at the output.  If so connect the speaker.
With no signal, do you hear much hum?  Does the idle current through the emitter resistors rise?

As Miyagi stated above, when the amp is running properly, delivering power to the load will increase the current through the output transistors.  This is normal, and not idle current, but working current.

If you connect generator to the FX return jack, can you get full power with a clean sine wave into a dummy load?
If so, what signal level is required from the generator to get there?

As far as the series or parallel regulators, I think parallel would be better.
It will reduce the current through the 12V reg., and possibly also reduce noise due to the fan current (which would appear at input of 9V reg.).

For your positive supply you have the neg. lug connected to ground, yes?
For the negative side you connect the positive lug to ground.

The floating/dancing/jitter on the scope sounds like a triggering issue.  The scope should have a few choices for trigger source.  For example, some have norm, auto, or line to choose from.  Try the different trigger options and see if you can get the waveform to lock.

The waveform you posted before did not have any crossover distortion, and that was with the trimmer all the way down.  None of the other pics have shown any crossover notch either, so I would just leave the trimmer set to minimum.
I think maybe your meter is not accurate at the low mV levels, and maybe it is biased hotter than what the meter is showing.  Do you have access to a better quality meter to try?
When you hear hum that gets worse as you turn up the bias, that is probably due to increased loading of the power supply.  Proper bias levels should not add any kind of hum.
Crossover distortion would only be noticeable at very low volumes, like on the decay of a note.  It would sound like a bit of fizzyness or graininess.
Other than that, any audible effect you are hearing from changing the bias is due to loading down the power supply and making more 120Hz ripple, which modulates the signal in different ways.  So you don't ever want to adjust bias in a solid state amp by ear.

Here is a picture of the crossover notch, which I don't think you have had at any time with your amp.

You cannot view this attachment.

Is that scope shot in post #206 supposed to be showing the crossover notch?  Because there is none there.

You are going to blow up your amp if you try to bias it for tone.
Stop replacing parts that you have not proven defective.

Post data everytime you make an adjustment.  Scope shot of the waveform, and DC readings across the .68 resistors with no signal.

Start by turning the bias all the way down and posting a pic of the scope showing the crossover notch so we can be sure we are talking about the same thing.
Get an output the same level as post #192 but turned the trimmer full CCW so you get the crossover notch.  Post the picture of the scope.

The MWC suffix part is also surface mount.  Any of the others will work.
The letters in the prefix designate the brand of manufacturer.

If you got the same pop at turn on by plugging it in with the switch on, that eliminates the switch as a cause for the pop.
Look elsewhere for a solution.

The -0.6V shown at the trimpot is if the trimpot is located as shown.
What did you find out earlier about the arrangement of the diodes and trimpot, as compared to schematic?  Why would you expect the voltage as shown on the schematic there to be correct?

It doesn't make sense that putting the 22R back increased the idle current, it should do the opposite.  The reason we removed it was because you could not get any idle current at all before.  So not sure what is going on there.

But that doesn't matter now.  If you were able to see a notch, and you turned the trimmer til the notch just disappeared, you are set to go.
15 to 18mV across 0.62 ohms translates to 25 to 30 mA of current, which is fine.

Play it.

You had around 15mV across the .68 with the trimmer set somewhere around mid-point I think.
That is a safe level.  Leave it there and proceed with the crossover notch info/questions that have been discussed.  Post scope shots.
The difference in voltage at the trim pot was probably due to the removal of the 22R across D8.  You don't need to worry about that, unless you find DC on the output, which would indicate a problem.

Those all seem good.
Where are you getting the voltages you were concerned about in post #177 ?
Have you tried scoping at the master volume pot lugs to see if signal is getting that far?
Use a gen. level that was giving you good output before.