A quick search found a 2015 paper titled "Time to let go? No automatic aesthetic preference for the golden ratio in art pictures."
The golden ratio is a frequently studied topic in many scientific disciplines, and, in psychology, it has been proposed as being a (universal) law governing aesthetic preferences. Empirical evidence for the golden ratio is equivocal and typically demonstrated through explicit (i.e., conscious, deliberate) evaluations using direct measurement methods (e.g., surveys). Here, we examined whether the golden ratio reflects an automatically elicited preference using the Implicit Association Test. We used real art images, with the foreground object presented in the golden ratio as well as either in the center (Studies 1 and 2) or ¾ ratio (Study 3). Both explicit and implicit evaluations did not reveal a clear preference for the golden ratio over other ratios. A possible preference for the golden ratio does not seem to be automatically elicited and may, rather, be driven by art expertise. This again calls into dispute the universality of a preference for the golden ratio.
So based on this it's disputable, at the very least, that the golden ratio is as pleasing as the received wisdom claims. I could not find something like a meta-analysis.