An image, done this morning, shows horizontal banding (perhaps difficult to see in the photo, but it's there) the banding is related to the image DPI and the image size. An increase in axis speed would make no difference to the dot positions although they would, of course, be placed at a faster rate (within the limitations of the laser operating system). The pulse repetition rate (PRF) must therefore increase with speed. As axis speed is increased there will come a point when the pulse width (PWM) is wider than the gap between successive pulses and no image can be produced. Perhaps not attainable with a moving axis but easily done with a scanner head. There must be an optimum axis speed but I think you are nearer to finding it than I.
Tweakie.