Should we have bunches of neutral body parts?

I think that it is something of a strawman.Firstly, the analogy to a moustrap is far from being entirely accurate.Secondly, Behe focusses exclusively on biochemistry so if we are following Behe's argument the only "parts" we should be looking at are individual gene-products (and ultimately genes.Thirdly, "new" body will be relatively rare and inobvious. We are more likely to see neutral modifications to an existing "part".Fourthly, and this further developes the third point, all but the simplest examples of Behe's IC are likely to come from coopting the workings of existing parts which evolved for other functions entirely. The idea that the whole structure must evolve in a single "go" for the current function is itself a strawman - and one the argument from IC relies on.

1) You really think that they eye evolved by parts just appearing ?2) You're using IC and a mousetrap as an example so I guess you are following Behe's argument. And Behe certainly didn't use the sort of anatomical examples you are trying to. Isn't it likely that Behe knows something you don't ?3) So the fact that the neutral trait is a modification to a functional (and therfore beneficial_ part suddenly makes the part non-functional ? No that would be a detrimental mutation.4) The eye isn't an IC system - the lens for instance is not necessary.
But given an example of the sort I am suggesting "chance" is misleading - evolution makes use of what is available. Statistically the chance that SOME useful coincidences will turn up is almost a certainty.

Your argument that we should have many neutral body parts relies on equating "trait" to "body part". Once you accept that the majority of traits are properties of body parts (in a very general sense) then it follows that "many neutral traits" does not mean "many neutral parts"As to the eye, a single celled organism can get use out of a light sensitive spot. It has no retina, no optic nerve and no brain. Now you may argue that that is not an eye, but evolution does not care about our labels and classification systems. It is clearly at least a possible - if distant - precursor of an eye.

For a single-celled organism to get the use of a light sensitive spot basically requires that it make a light sensitive chemical (and that's pretty easy) and that the chemical reaction that occurs when light hits it influences the organism's behaviour in a useful way (since the behaviour ultimately is chemical the only difficult bit, really is the "useful" part). The behaviour can be as simple as moving toward or awat from the light. Localising the chemical in a single spot can be left for later evolution (the benefit of that is more precision in locating light sources).

That depends on what you count as a retina, and what you count as an eye. Light sensitivity almost certainly has to be one of the the first steps since the rest seems to make no sense without it. The lens on the other hand is very likely a relatively late development because even now nautiloids have eyes without lenses that focus on the same principle as a "pin-hole" camera.Now I don't know of any resources off hand on how the retina developed - and it would depend on how multicellular life evolved. Which we really don't knwo a lot about yet, although research is discovering more On the view that multicellular life started as colonial single-celled life and that there followed a period of growing specialisation and integration it might be that the first step was the first stages towards a nervous system - a communication channel. That step allows some cells to specialise in light-detection while others abandon it to devote their resources to other functions.Now even bacteria (or at least some species) can communicate to some extent by chemical means, so maybe that developed seperately and then came together with light sensitivity - or even maybe intercellular communications came first and light sensitivity worked through that. But it's hard to call chemical communications between cells any sort of eye at all.