I don't know where rods and cones come from. But what's more likely: They're specialized from variations on precursor cells, or they came with full instructions delivered by a virus attached to a comet, originating somewhere outside of the known laws of the universe?
It is like an unsolved murder. We don't know 'who dunnit'. The cops have a perp captured who I don't think did it. You are the detective who thinks he DID do it. We both look at the same evidence but we interpret it differently.
I agree that small variations on pre-cursors can happen - probably do sometimes, here and there. I do not think RM/NS can happen on a small number of nucleotides at a time leading up to a long nucleotide sequence change. For me, the incredulity springs from doing a little arithmetic on the amount of possible permutations of a small sequence. I did this for myself years ago. In order to pick a sufficiently small sequence to make the amount of permutations small enough that it is credible that most of the combinations of these sequences readily occur with random mutation, those size sequences are very short. On the other extreme, picking the size of sequences such that it is more credible that they are long enough to code for something differentiable, they are so long that there is not enough time for a significant fraction of the combinations to have had a chance. It is like looking for "Leave it to Beaver" in a random sequence.
In your view, is there a difference between an intermediate form and a not-intermediate form?
Let us say there is some protein that is a presumed pre-cursor to another protein. Let us say that there are minimally 256 bits of changes that have to be made along the path to the target. The synthesis of the target is presumed to take N steps, in 'evolving' from the precursor to the target protein. N could be 256 steps, one bit at a time; N could be one step, in one miraculous move; N could be more than 256 steps because of a circuitous path with meandering inefficient scrambling. I would think that the best approach to a credible proposition is that N is somewhere between 1 and say, 16. Each of these steps would have to be intermediate forms, defined by: they each provide useful function to the organism. So if one can theorize such a path, of on average 16 bits change at a time, each time producing a useful function and not losing the needed function that the pre-cursor provided, and do it 16 times, then that would be a model that could show how intermediate forms work in such an example.
If N=1 then the 256 bit change is one out of 10^77. That's why I called that a miracle. A 16 bit change is a manageable one out of ~10^5. What stretches my credulity is that there could be found 16 intermediate steps that could happen 'just so'. This is such an immense problem to prove. In the environment it is reasonable to say that most the intermediate proteins are gone or they are too hard to find. The amount of computation that would be required to 'find' intermediate proteins that fit the criteria would exceed what could be done in the lifetime of the universe, much less the lifetime of a human. So we can only wave our hands and say 'it could happen', or we must doubt. I doubt.
There is no "neutral" problem.
In presenting above the changes to the pre-cursor molecule in steps toward the target model, do you not agree that some steps could be 'innocuous' but yield no extra function, and so that step is neutrally affected by NS because the yield provides no advantage at that step (but may in the future?)?
Don't misquote me. I said "re-use appears to be a common 'strategy' in evolution." I don't know why you get so hung up on this. We talk about survival "strategies" as short-hand for the attributes of a particular organism that increase the likelihood that it will reproduce.
Why I get as you say 'so hung up': If there is no teleology to evolution, which is
a priori to the theory (agree?), then why cannot an evolution-enthusiast express anything without teleological terms. If it is mere anthropomorphizing, then challenge yourself to re-state without such terms. Can you do it? If you cannot, my point is that maybe you are blind to the notion that you, on some other level of thinking, DO attribute purpose. Here in this forum of skeptics, is it not heresy to attribute purpose? Your 'short hand' should be avoided if it always leads you to appear to attribute purpose when your creed is 'there is no purpose'.
I think the best you can do is just say: that the mechanism of evolution appears teleological but it is not. The reason it is not is because we believe it is not. Disclaimer: any terms we use that imply purpose are merely shorthand and it is not intended to attribute 'purpose' to how evolution works.
Translation: it is impossible for us to describe what we believe without using words that represent what we do NOT believe.
Again, I'm pointing out "cognitive dissonance".