In the study, participants with no background in art, were shown an original image and distorted versions of a statue. The original statue’s proportions reflected the golden ratio. The original image strongly activated sets of brain cells that the distorted images did not, suggesting beauty is partly an innate quality.
This doesn’t sound like the best study design unless they also tested with statues whose original proportions were not in the golden ratio. Otherwise you’ve just proven that people can tell when a human statue has been distorted and don’t like it.
I’m also unclear as to exactly how the vertical and horizontal lines on the Mona Lisa were chosen. Why didn’t a1 go to the top of the head? Why did a2 go up to the eye and not some other facial feature? What determined the endpoints of b2? It seems like there were so many options available and no natural way to select between them, so I have to assume the endpoints were selected simply to produce a golden ratio. And one could do that with any image at all.