This bed was not designed for cats. It's not a cat bed. They've just made use of an object that was already there. Turns out, evolutionary biology is chock full of this kind of thing.
In fact, there's even a word for it — spandrel. Coined by Steven Jay Gould, spandrels are phenotypic features that are useful to a plant or animal even if the feature didn't specifically evolve for that use. Basically, it's the evolutionary biology of accident. Not all of our features were specifically sought after by evolution. Some , instead, are the side effects of selection pressures directed at different parts of the body. In other cases, selection pressures shape a feature, and we end up using that feature in ways that have nothing to do with why it originally evolved. Spandrels are the antidote to overly-tidy "just so stories" about evolution and they're important to keep in mind when you read headlines that proclaim, for instance, that the human hand evolved for punching and the human head evolved to get punched.
Brian Switek explains why spandrels matter, and why the head-punching paper is a particularly egregious attempt to ignore the role spandrels play in life.
Although Morgan and Carrier focused on the bludgeoning qualities of modern human hands in their previous paper, their new review suggests that our ancient relatives and forebears – the australopithecines – had faces that were molded into punching bags by natural selection. No sooner did humans come out of the trees, Morgan and Carrier suggest, than they started whaling away on each other. The trouble is that they undercut their own hypothesis, leaving only a crumpled heap of speculation.
Citing crime statistics from western countries, Morgan and Carrier write that fistfights often result in broken noses, jaws, and other facial bones. Therefore, they reason circularly, prehistoric humans that punched each other in the face should have more robust facial bones to cope with such blows. Given that early humans Australopithecus and Paranthropus – the latter often called “robust australopithecines” – had broad faces with wide cheeks and thick brow ridges, they’re obviously perfect candidates for Morgan and Carrier’s favored interpretation.
Morgan and Carrier didn’t study whether or not the hands of the early australopithecines could form a fist. Their previous work was on our species, Homo sapiens. Nor did they look for signs of broken facial bones or blunt-force trauma on prehistoric skulls, or even try to model how early human skulls would have reacted to the stresses of an incoming fist. The entire argument is simply that australopithecine skulls look like they could take a punch.