However, I would argue that these are precisely the wrong questions to be asking.
There is no doubt that humans have a certain concept of what is considered beautiful, but it is hardly adaptive. In fact, one would be hard pressed to claim that it is much more than an abstraction. Anyone spending more than a few minutes in public can't help but be struck by the incongruity of our supposed standards of beauty and the actual pairings that occur. There is little doubt that most of us are not Brad Pitt or Angelina Jolie [note my own arbitrary notions of attractiveness].
Certainly we would recognize the difference between youth versus middle and old age, but this still doesn't fit the overall concept of beauty; especially not as an adaptative trait. After all, the one element we don't find, is a selection pressure to increase beauty. We rarely find individuals that don't form long-term relationships and have children based on a criteria of beauty. So what gives?
Part of the problem is that beauty is often measured from photographs, which should be the first hint of something gone wrong. Human vision and assessment of individuals has NEVER occurred based on such instantaneous moments. Our concept of an individual's appearance is a continuum of observing them in all manner of motions and situations. So, I would argue that any image that is "frozen in time" is simply an artifact that we can study, but it is purely an abstraction and not anything that would relate to our adaptive sense of beauty.
In fact, there is a strong reason to believe the old adage that "beauty is only skin deep", which suggests that our human concept of beauty and attractiveness entails far more than simplistic facial symmetry.
First, let me acknowledge that we all have a sense of what we find attractive and beautiful. In fact, there are many instances of where we can consider someone as beautiful without actually being attractive to us. As a result, my conclusion is that this is simply an abstraction in our brain imaging. Just as we can look at a landscape and find it beautiful, so can we look at another individual. In fact, we even do it with animals. Whatever the mechanisms are that allow for such judgements, they are specific, but they don't necessarily lead to anything more.
Similarly, we need to distinguish between simple sexual appeal versus longer-term relationships. While it is tempting to argue that sexual behavior is adaptive, most people can readily admit the results of intoxication [or other inhibition reducing states] and how readily our "standards of beauty" are compromised simply because we want sex.
However, biology is about mating, not sex.
From this we can reason that beauty is certainly a precursor to attractiveness and ultimately may lead to a longer-term commitment in a relationship. However, none of this addresses what our sense of beauty actually translates into a long-term relationship.
In truth, beauty is much more complicated than mere physical appearance. It also includes our physical perception of this individual as they move and interact. Behavior clearly plays a significant role and makes a huge difference in our ultimate assessment of initial beauty. In other words, someone may appear beautiful to us until their behavior becomes offensive, in which case we generally begin to find them less attractive and less beautiful. In fact, this is a significant part of the appeal in photographs ranging from glamour magazines to pornography. The photograph can act as a proxy for actual interaction and as a result becomes a "silent partner" in allowing the brain to construct whatever fantasy is most appealing.
One can get a glimpse of this distinction by watching performers where their behavior in one context [i.e. movie, concert performance] can elicit a particular sense of their attractiveness and appeal, while seeing them as normal people can offer a radically different perspective.
In all these cases, I would suggest that there is a completely different mechanism at work. The process of assessing something as attractive comes down to our own imaginations and experience. Events in our lives that made us feel a certain way, positive or negative, will become factors in assessing the attractiveness of someone else. Whether it be a particular look, or a particular motion, or the fantasy that it plays into ... these will all be factors in determining how appealing [and attractive] we find the individual in question.
Again, this is all too readily apparent when we encounter the conflict between a performer that we like and their real-life persona. Our perception of the latter will then influence our feelings [including sense of attractiveness] to the former.
While I have no doubt that many of the studies performed are exactly correct when individuals are forced to make choices based on their perception of beauty or sexiness or health, etc. (1) What is missing from these studies is to follow-up and determine who these individuals actually chose in real life to pair with. That is the element that represents our adaptiveness. In addition, I know of no particular studies that assessed the changes in attractiveness that occur when viewing the behaviors of the individual in question. Without that information, the conclusions are simply window-dressing.
(1) What is most surprising are studies that ask men to assess reproductive potential. In my view this represents an arbitrary value assigned by the researcher and not something that is generally intrinsic in men themselves. Again, this is readily demonstrated by the number of infertile couples that somehow failed to express this "adaptive" quality. This clearly demonstrates that neither men nor women are very good at assessing the viability of reproductive success, so it is difficult to argue that this is an evolutionary result.