The only one you need details to is the aligned-holes experiment. The others are just observation. You watch the bottom of a ship vanish before the top. You watch a swinging pendulum deviate from its beginning alignment more and more over time.
Also, ships do not always sink below the horizon. The right atmospheric conditions can refract the light in a way that causes them to be visible at a much greater distance than the straight geometry would allow. But this explanation can be inverted to claim that it is the vanishing that is due to atmospheric effects and that the non-vanishing ship is the accurate reflection of the geometry. Falsifying that hypothesis is, again, not so easy.
(Presumably this is also true at sunrise, so the day is longer than it "should" be from purely geometric considerations.)