"Drastically misaimed" really says nothing about whether or not a cuteness instinct would be a good adaptation, though. A counterexample: it's a fact that our visual systems are acutely sensitive to rapidly-moving things. The evo-bio hypothesis is that this is predator detection. Does the fact that 99.999999% of the rapidly-moving things I notice aren't predators negate this hypothesis as well?

I can't think of very many cases in which people endanger themselves or their reproductive chances for the sake of cute animals. I'm sure it's happened once or twice, but using this argument means demonstrating that the number of potential children lost due to finding bunnies cute is greater than the number of actual children attended to due to finding them cute.

As an aside, I think that Google in this case is adding to the confusion. The evo-bio cuteness theory is generally stated as being about a system that detects facial markers that strongly differentiate babies from adults - the key ones being eyes large relative to head size, pursed mouths, round cheeks, and round chins. Some baby animals, when viewed up close in Google, display some of these characteristics. In the wild, however, baby animals are almost never seen up close, and even when they are, they trigger the facial recognition systems only in dribs and drabs, like bad CG.