Beauty and the Brain illustrates research showing a correlation between the closeness of a stimulus to a prototype or stereotype and a judgement of attractiveness. The research attributes the correlation to the ease of processing. The quicker we "recognize" the "classification" of something, the more likely we are to find it attractive. In other words, faster abstracting tends to give us a more positive recognition factor. Does this sound like we are more prepared to have signal reactions than symbol responses (proportionally)?

The report stated, Even more significant, Winkielman said, is that when processing ease was controlled – when, that is, the categorization speed was factored out of the equation – much of the relationship between closeness to prototype and attractiveness disappeared.

If the correlation mostly disappeared when processing ease was factored out, it would support the claim that processing ease or quickness of recognition dominates "closeness to prototype".

I can see that "identifying" an abstraction to our current map more rapidly would have have survival value. We only need to recognize a tiger, not how close to the prototype a tiger stimulus is, in order to run away to survive. The faster the recognition, the greater the survival probability becomes.

I don't understand how "It seems far more likely that we like the one closer to the ideal because it comes closer to the ideal." has any explanatory value.

we like the one that comes closer to the ideal because it is the one that comes closer to the ideal.

We like X because it is X?

I suspect that the survival value of rapidly recognizing something gets instantiated in the nervous system as a pleasurable - reward-like experience, and that is the mechanism that translatets into perception of "beauty". We "like" it more because it "feels" more pleasurable, and that's just evolution's way of rewarding learning. See The Hedonic Response in Problem Solving for some additional abstractions. "We like it more when it's easier to recognize" becomes a general functioning bias as a side-effect of the hedonic response in problem solving.