Great work lately. I've been following your 'throw the fat man' series with enjoyment as I've tried to make similar claims for years that impulsive moral decision-making should not be expected to be a test of purely abstract reasoning, and that the fat-man test in particular doesn't test whether people really make large moral distinctions between passive and active efforts.
My view has always been that the facts of the scenario tend to stimulate instinctive or culturally-conditioned reflexive responses that generate subliminal hesitations and tribal preferences.
So, when the examiner says 'throw the big fat man off the bridge', he thinks that stands for the abstract example of 'kill a random human, that we assume is sure to stop the train', but the examinee feels, just below the state of conscious awareness, 'Push a big fat man?! That's really dangerous! I might fail, and he pushes me instead, without even saving the trolly - that's the worst of all worlds!'
Since the 'moral function' often manifests itself in the form of post-facto rationalizations for reptilian-calculations of self-interest, it's not hard to get people to try and piece together a string of words which 'justifies' their failure to intervene on the basis of asserted 'moral principles'.
But as with you, the typical response of people who really want to believe in the meaningfulness of this scenario is that either I, or these people, are just unable to think abstractly, or are making ridiculous 'not real enough' criticisms of an abstract test.
One skill tested by IQ tests is the ability to de-contextualize, to move from the concrete to the abstract. But the older I get, the more interested I am in re-contextualizing, moving from the abstract back to the concrete. History is an old man's game.
I noticed that analogy to Newton's laws on your site, but of course, the analogy is false. Newtonian mechanics posits idealized assumptions to draw conclusions about abstract natural laws. But experiments in moral psychology are making abstract assumptions about the workings of the human mind, which are not like natural laws of Physics. They are putting the cart before the horse and assuming away psychological tendencies that have major impacts on the responses of the study participants.
So, to test my theory, I have often proposed to people who quote the 'throw-the-fatty' examples at me, that they rephrase the question with different characteristics of the individual to be thrown.
So, rerun the test with 'throw the old fat woman', or 'throw the fat black child', or 'throw the wheelchair-bound invalid', or 'throw your favorite celebrity', or 'throw Obama', or 'throw an Asian communist'. For progressives, or open-borders advocates, I would suggest 'throw Steve Sailer' - who happens to be a large male like in the original scenario - but I imagine that one would get a great deal of enthusiasm for active intervention and explanation of moral justification.
This has proven persuasive to my interlocutors on several occasions. And it's made me permanently skeptical of results from the field if they don't apply some kind of control or test for the impact of these sorts of variations.