OK, maybe you missed it. This is what that metanalysis you cited said about average faces:
These results don’t mean that all attractive faces are average…or that average faces are optimally attractive.
In other words, average faces (composite or not) are more attractive on average. Go ahead and actually read the study if there’s anything you still find “exceptionally hard to believe.”
Don’t know how to make this any clearer. The study itself concedes that not all attractive faces are average and average faces aren’t necessarily the most attractive. That concession is exactly the point I’m arguing.
It’s evolutionary psychology aiming to study what people find attractive and why. The main hypotheses considered in this metanalysis are:
- People find certain faces attractive because features of those faces are genetically linked with robust health or similar.
- People find certain faces attractive because the brain processes information in a certain way (e.g. we might find symmetrical faces more attractive because our brains are better at processing symmetrical images).
It concludes that there’s more work to be done, especially on the first hypothesis, but that both are likely to play a role in what people find attractive.
Edit to add: One of the factors examined in what people find attractive about faces is “averageness”. This analysis concludes that “average faces are more attractive than most faces” while conceding the point that I quoted above.