Hello,
I have a QHY5 camera and there's a ST4 option which I tried as well ("on camera", bypassing the ASCOM). Actually, that would be my second question as this failed as well. It seems with ST4 cable doing a "manual guide" really does something for each direction as I can hear a distinct sound coming from gears for all 4 directions, however when looking at the star, combined with its own (earth's) movement, it's really hard to tell whether it moved in all 4 orthogonal directions, for example it's possible it moved "south" two times, when clicking "south" and "west" or perhaps not moved at all because of massive backslash even though I was able to hear the sound. Calibration result is the same as with eqmod with orthogonality error being at 70-80, nowhere near being perpendicular. Now that you say that slew and pulseguide commands differ (i.e. not just slow vs. fast), I'm seriously thinking something might be wrong with the mount (bought a used one), as if it's interpreting pulseguides incorrectly and either ignoring RA guides when using eqmod or moving either in a wrong direction or not moving properly when using ST4. Sadly, problem is very specify and hours of googling didn't help.
I tried drift align nevertheless accepting the broken calibration and to my surprise it actually said my polar error is rather low (azimuth adjustment step), especially when considering I'm doing this from my balcony window and from where I'm not seeing polaris (used compas, did some adjustments by looking at the moon). It was actually tracking the star very well (didn't even try altitude adjustment step yet), but if you look at the attached screenshot (using ST4, but looks the same with eqmod), you'll see something is very wrong with the graph (btw., selected to show pixels rather than default arc-sec in settings because I couldn't see the red line with arc-sec selected). Also, as soon as I stopped drifting and started guiding, it expectedly quickly went off and bailed out because it was unable track the star.
Thanks,
Vedran