Couple of questions and a comment.
One, why can't I get the print at w140 ? I went to that Polish site and got it and it is scanned terribly and there are tings missing. Then I went to BAMA and it appears to be the same file but also on the list was "missing pages". None of them seem to have the end of the front end. Where is a good manual for this ? I don't see any TP1 or 2, the PDF I have is 47.4 MB. Is that the same as you got ? If so, can you tell what pages those test points are ? Might help.
Two, your pictures show 2 waveforms, one severely overpeaked and the other severely underpeaked. What is the difference in the settings between the two ? Looks like the attenuator is set the same and both CH 1, are they at different frequencies ?
Now, one of my general rules about troubleshooting is that when it doesn't work right look at the last thing done to it. you did the multiplier which should have no bearing on the peaking. you changed the vertical output board. not if both channels are affected the same way that means the problem is there. If it is different between CH 1 and CH 2 that is a different story. But you only changed the output right ? If both channels display the same there is one thing they have in common.
Both David and Fabio are right, the 5 mV range is where to start. If it displays the same at that setting then get out of the attenuator. (which I call the front end) Forget any adjustments there.
If you find the display the same on both channels down there, whatever compensation is on the output board is the most likely culprit except for the outputs themselves.
Vertical deflection plates on a scope can affect focus, but that means there is an imbalance in the output stage. When you stick a probe on there you just put 10 meg to ground. (I hope you use 10 X except in rare cases when you need all the sensitivity) If that causes a DC offset of only one plate, or a common mode offset in both plates that could be the cause of the problem as well as the compensation problem. But to determine that we need the settings used on your 2 quite different waveforms. there are o clues, no readout on the screen and we can't see the time base, all I think I saw (pretty sure) is that both waveforms are from Channel 1.
If by chance is it intermittent, then it is an intermittent cap that goes shorted and then open when it feels like. Not the most uncommon failure mode in the west. That's where the frequency might help to determine the cause, the higher pF value caps seem a bit more likely to do this. And when it comes to frequency comp they are not using tantalum.
I hope this is of at least some help, so I put it out there. If not just ignore it. but that's the way I would proceed, go to 5 mV and see what that looks like, both channels. And don't try to align it anymore. That is far too much error to be simply in need of alignment.
Jeff Urban