So it is a case of draw the conclusion first from the outset 'the Earth is flat' and now make whatever you observe fit in with that presupposition whatever it takes?
That is precisely my problem with the invocation of zetetic methods in support of flat Earth theory. It seems if you make ten zetetic observations, the one that "the Earth looks pretty flat" is given priority over all others and you have to disregard those or invent magical theories for them. Why wouldn't you assume that the other nine that have obvious explanations are the ones to base your model on and that the Earth looks flat is the illusion (or, more accurately, exactly what you expect from an observer with a low POV on a huge sphere).
For example, if I did this experiment, I would assume that the two objects were being illuminated by the same light source.
But some people will argue that the moon is emitting it's own light and it's just a big coincidence that it's phases always match the golf ball. It's also a coincidence that the stars rotating around on the dome at the North and South Celestial Poles exactly match what would be expected by observers on a spinning globe Earth, and so on and on and on.