If the intuition fails on simple case X, the next thing is to try to play to see why it fails in hopes of conjecturing an assumption that rules out all the failures; likewise one can try some other simple cases in hopes of noticing the pattern. I think your example of the leap complexity paper isn’t a good one because the authors had a correct statement—what I would’ve preferred was an example where someone was technically incorrect (especially if deeply incorrect) but “morally correct” in a way that’s vindicated when the correct theory is figured out.
If the intuition fails on simple case X, the next thing is to try to play to see why it fails in hopes of conjecturing an assumption that rules out all the failures; likewise one can try some other simple cases in hopes of noticing the pattern. I think your example of the leap complexity paper isn’t a good one because the authors had a correct statement—what I would’ve preferred was an example where someone was technically incorrect (especially if deeply incorrect) but “morally correct” in a way that’s vindicated when the correct theory is figured out.