Your observations have true base and the Vpp or Vrms will be the best for comparative results, but here we have the same methodology, the same position, the same time, the same probe, the same vertical resolution (50mV) the same or very closed horizontal time (400/500usec) and the closer memory buffer 1-2M).
Additional from what I see there is stable optical depth of camera.
I think that they are comparable.
and all that didn't matter when you do visual compare because of different display size, resolution, pixel per DIV, DIV per full ADC scale and finally dot size. In case of Owon and Hantek the resolution, size and DIV per ful ADC are different, but luckily Owon is having exact the same display type(tech. type)/manufacturer, so the dot size is the same so one can take screenshot from Hantek and 1:1 paste on Owon screenshot to compare (where of course one need to know that Hantek is doing as well 10DIV vertical, but visible are only 4.2 DIV where Owon is doing 10DIV and visible 10DIV). Siglent can't be compared to them as there are major differences.
You need to know as well that the way how the waveform is being displayed matter as well, this is where it make sense for visual compare to set intensity on full. The horizontal resolution (or actually visible dots per div and the way how they will get displayed) matter as well, there is difference between DSOs displaying only some peak dots or some amount of dots or all dots.
Anyway, let's compare measured values ...
Mine after all modifications to this test with the BNC Shorting Cap has 12-16mV (10X) on CH1 and 14-18mV on CH2.
... from stock Hantek DSO5102BMV (50mV/DIV, 80us/DIV, 2Mpoint depth, trigger on peak to catch all spikes).
CH1 does have some more noise injected from the display, without display the level is the same as on CH2.
Question1: did you measured in peak detect trigger mode?
Question2: i assume the Vp displayed on your picture is in real Vpp and not real Vp, isn't?