Every time I've looked at the horizon I was seeing half land and half sky, with an apparent line cutting through my vision. Take a mirror and turn around, with the horizon facing your back and, when studying the mirror, the horizon follows the level of your eye: See
Experiment 10 of Earth Not a Globe.
The default is, therefore that the horizon is at eye level. If someone has a crazy theory about the earth being a ball and the horizon being imperceptibly below the eye level, in contradiction to observation, it seems that the onus is on that person to demonstrate their claim.
None of this is to say that the horizon is always at eye level at every altitude and atmospheric condition, or that one could expect to see the same from a plane where the horizon is very foggy, just that it has been tested at sea level to be so, just as Rowbotham tested it from the third story building in
Experiment 15.
I understand that the common response to this is "that's too low," or whatever, but yet, it remains a test of the horizon. If one is to argue something about the imperceptible drop, that is a claim against reality, and thus one should be burdened to show it rather than argue that it is the burden of others to prove that there is no imperceptible drop.
Asking others to prove that there is no imperceptible drop is akin to, in a discussion about ghosts, asking someone to prove that there are no ghosts that they cannot see. Surely we can see the fallacy with that.
Rowbotham tests the matter further by testing the altitude of distant bodies on opposite horizons, and they are lined up, providing additional evidence that the horizon is at eye level. This is a better test than a direct test of the horizon, since one can always argue that there is fog in the distance that just seems sharp because of its distance.