Humans have been designing cameras for a century, and telescopes for several centuries. So, optics is well understood.
As we go down the eye sequence, we want visual acuity to be improving. Nilsson and Pelger measured the acuity by calculating spatial resolution.
(The basic idea is that each light sensitive cell really should have a slightly different field of view from any other cell. When that is true, then different cells receive different information. The whole eye gathers the maximum possible amount of information.)
Resolution is not the only issue. Things such as the size of the eye and the brightness of the scene also affect acuity. However, Nilsson and Pelger's equations take care of all this. For each step in the sequence, the equations say that the eye has improved. And, there is a smooth gradation between each step.
Specifically, Nilsson and Pelger broke the sequence down into 1% changes. What I mean by a 1% change is that a length (or whatever) got 1% larger or smaller. When you look at it that way, the whole sequence, from eye spot to fish eye, turns out to take 1829 1% changes.
Nilsson and Pelger have a graph which shows the logarithm of the resolution versus the number of 1% changes. The graph is a straight line. This means that there aren't any inefficient areas in the sequence, where a lot of change will cause only a little improvement. It also means that there aren't any areas in the sequence where change has no benefit, or has negative benefit. The graph is proof that the sequence is, indeed, a Darwinian scenario.