I've decided to upvote your article for presenting a concise summary of this topic and also, very importantly, including links to the full pdfs of the academic articles Weisberg et al. (2008) and McCabe & Castel (2008) which you used to support your argument. This is not always possible, but I think is always to be commended when it is.

However, having now read both of these articles in their entirety, I would like to point out what I believe to be an important omission in your summary of the Weisberg et al. (2008) article.

First, you state:

And yet, Yale cognitive science students rated the 'with neuroscience' explanations as more satisfying than the regular explanations.

You also state:

Somehow I suspect people who chose to study cognition as information processing are less likely than average to believe the mind runs on magic.

You conclude with:

Sometimes even physicalists need to be reminded — with concrete reductionistic details — that they are physicalists.

However, I obtain an entirely different conclusion from this article. First, I shall summarize what I believe to be the salient points of Weisberg et al. (2008):

• Novices (ie. those who had not chosen to study neuroscience), found good explanations satisfying and bad explanations to be unimpressive. Add useless neuroscience and what happens? Good explanations remained pretty much just as good but bad explanations suddenly looked a whole lot better. As stated in the article:

Post hoc tests revealed that although the ratings for good explanations were not different without neuroscience (M = 0.86, SE = 0.11) than with neuroscience (M = 0.90, SE = 0.16), ratings for bad explanations were significantly lower for explanations without neuroscience (M = −0.73, SE = 0.14) than explanations with neuroscience (M = 0.16, SE = 0.16)

• Sadly, students who chose to study neuroscience were only impressed if given explanations actually containing neuroscience, despite the fact that these additional frills were worthless:

Unlike the novices, the students judged that both good explanations and bad explanations were significantly more satisfying when they contained neuroscience, but the bad explanations were judged to have improved more dramatically, based on a comparison of the differences in ratings between explanations with and without neuroscience [t(21) = 2.98, p < .01]

• Thankfully, neuroscience experts were not fooled. Bad explanations were still rated as bad. Also, the good explanations were rated as worse when the neuroscience information was included:

Good explanations with neuroscience (M = −0.22, SE = 0.21) were rated as significantly less satisfying than good explanations without neuroscience [M = 0.41, SE = 0.13; F(1, 46) = 8.5, p < .01]. There was no change in ratings for the bad explanations (without neuroscience M = −1.07, SE = 0.19; with neuroscience M = −0.87, SE = 0.21).

I think the problem is not that the neuroscience students needed to be reminded that the mind does not run on magic. I would instead say that they give too much value to the explanatory power of neuroscience, even when superior explanations exist. I do not think this is surprising given that these are people who are: a.) fairly untrained and b.) have chosen to study neuroscience, and therefore are more likely than the general population to have a highly favorable impression of it.

The conclusion I would draw from this article is that irrelevant neuroscience information should generally not be included unless you feel your argument is shoddy and your main intent is to convince novices that you are right. Only neuroscience students are more impressed with the addition of useless neuroscience in any explanation, and they are presumably not a large portion of the general population. I am certainly not confident that this result can be generalized to anyone with any passing interest in neuroscience, or anyone who wants to be a rationalist. Also, neuroscience experts will really not be impressed with the addition of such irrelevant frills. This says to me that the best way to promote the proper understanding of the way our minds work is to never include neuroscience information unless it logically adds value to an argument and to train students in specialized fields like neuroscience to value logic above the appearance of something being about neuroscience. Apparently, this happens already, since experts seem to hate the useless neuroscience information so badly that it taints even the good explanations in their eyes.

In regards to McCabe & Castel (2008), I wonder if people simply feel the concrete brain image is more comprehensible to them that the abstract graphs (both the simple and complex ones). It may even be that it takes them less time to understand the intent of the brain scan images because of this, therefore they like them better. I wonder if other images that also concretely demonstrate an effect would be rated higher than ones that don't. Also, perhaps the brain information is preferred because it conveys more information than the bar graph and but is also easier to grasp than the complex graph.