I don't think I would agree with the claim that "natural, biological systems are actually often perfect models for ... efficiency." Natural, biological systems tend to get the job done (natural selection at work), but often do so in bizare, highly inefficient ways.
For example, most of us have eyes. Our eyes generally do an extraordinarily good job absorbing reflected light and allow us to perceive an enormous amount of visual information regarding our surroundings. So far, so good.
Look a little deeper, though, and the structure of our eyeballs quickly shows the vestiges of it's bogosort design process: vertibrate eyes all have a blind spot where the optic nerve blocks some incoming light from reaching our photoreceptor cells. We generally don't notice this because we have two eyes, and our brains are pretty good at merging the images we get from each one to cover for whatever the other missed (including constructing some outright fabrications where needed). Essentially, the human eye is a camera with the power cord routed across the lens: an obviously idiotic design decision that persists because it wasn't quite bad enough to be completely debilitating and could be mostly compensated for. Cephalopod eyeballs, which evolved independently of ours, do not have this particular weakness (although they do have their own suboptimal quirks).
It's not hard to look at the bevy of ingenious yet plainly stupid constructs that evolution has created and decide that they fall well short of any idealized standard of "perfection." Why should the Borg accept a visual sensor with such a glaring flaw, when they know they can do it better?