Our decision-making under the influence of science comes back to the question of how much we should trust models. There is no easy general answer. But among the many ways this demonstration is interesting, it makes it very clear that some models are extremely trustworthy. On the other hand, obviously some models are useless. So there must be a continuum.