The "curve" most everyone seems to want to look for to distinguish a globe from a flat earth is (I believe) a false quest. Here's why:
When looking from high above either a sphere or a disc, the "horizon" of either transcribes the base of a cone. If the viewpoint is the apex, the angle from horizontal in all directions is a 2D circle, either the circumference of a sphere (or spherical cap) or circumference of a disc (or some smaller circle on that disc).
From the surface of the base, the lateral line is "flat" but it surrounds you in all directions. It "curves" toward you radially as you turn 360°. As you increase height over the base, it'll keep doing that, but you'll start to see the "curve" of that arc of the circle. It's not the "dip" of the horizon left or right because the angle looking down doesn't change.
Even when you can see the whole of a flat disc or whole of a globe, you still don't know which is a disc and which is a sphere, unless you can determine relief in the z-axis (the same axis as height) or you can gain an oblique angle, in which case the 3D disc will reveal itself while the sphere will be retain symmetry.
The "curve" that matters in distinguishing sphere from flat isn't along the transverse of a horizon. It's the "dip" toward and away from the viewer, not left/right.
The right honorable Samuel Rowbotham didn't understand this either when he incorrectly presumed this should be an attribute of a rotund earth in Experiment 7 of "Earth not a Globe"
Likewise, globe defenders claiming to see the horizon curve from high altitude aren't actually resolving the dispute either. The horizon being straight (or not straight) across the horizontal is not a marker of globe/flat.
Is there anything wrong with that contrary argument?