>Plus there's the fact that the human brain is set up to comprehend things by forming an image of A acting on B, rather than of some abstract action by abstract agents taking place in the ether. So a reader would have to do considerably less processing to comprehend this paragraph in active voice.
I don't disagree that this particular example could be improved as you say, but your generalization from that seems suspect. It's entirely dependent on whether the agent or the patient is the point of interest.
(I've noticed the passive voice is strongly discouraged by American style guides, even in cases where it is clearly the better choice, whereas it's more widely accepted in the UK for some reason. Given this, it's possible that familiarity has some effect on the brain's ability to comprehend.)