Actually, it makes perfect sense for sexual selection on sexual-attractiveness-features to be subject to far greater selection pressure and fine-tuning than baby-cuteness.

I'll make a testable prediction here: Cases of parental superstimulus (like baby ducks following a stick figure, infant monkeys getting attached to puppets, etc., if I'm remembering correctly) ought to be far more common / easier to fake than sexual superstimulus. I'll limit the key part of the prediction to complex vertebrates so that they have large enough brains to be complicated, but I wouldn't be surprised to find the rule more universal than that.