Hi Annafan,
I had been meaning to reply to this when it got promoted, but I've been a bit distracted.
I think you make a good case in general, I'm not sure of the utility of producing an expanded list of examples of redundancy and robustness since so often the ID/creationist crowd will consider just one example of what they consider an irreducibly complex system to be sufficient proof of some extranatural component. I also agree that the 'junk DNA' point is fairly irrelevant and perhaps confusing, 'junk DNA' is a fuzzily defined term at the best of time, if we consider superfluous 'junk DNA' to be that which is non-functional then we might just as well include it in the set of all effectively non-functional DNA, although this isn't a set we can readily identify either in coding or non-coding sequences.
I'll try and address some of your specific questions.
gene duplication errors create redundant copies of genes. Those copies can take over if the original one is no longer functional due to a bad mutation (q:is it a given that multiple copies will ALWAYS only have positive effect?)
No, it is possible that over expression of a gene by a duplication would be detrimental, although it will vary greatly on a gene by gene basis if a simple duplication would produce a large enough dose for a detrimental response. It could also be argued that as well as adding a layer of redundancy a duplication could increase the chances of a detrimental dominant negative mutation which would block the effect of both copies of the gene.
sexually reproducing organisms can make use of the "heterozygote advantage" because they have all chromosomes/genes duplicate (except X/Y of course). One copy can be dysfunctional without fatal consequences (q:is it categorically NO effect if one is dysfunctional, or is there a whole spectrum from "NO consequence" to "not deadly but still seriously affected"?)
There are many cases where even a totally null allele is no only not fatal but possibly not even phenotypically detectable. n regard to your question; as with duplication there is a possibility of a dose effect and there are known examples of where a low dose causes problems, termed haploinsufficiency.
an example like vitamin C illustrates that often it is possible to take in certain essential substances (like, via food) instead of letting the DNA synthesize them internally. This is why humans and apes (and bats) have a defective gene for vitamin C synthesis without being affected by it. (q:is this a fairly unique (known) example, or is this abundant?
I really don't know about this, I can't think of another example off the top of my head.
biochemistry with complex biochemical molecules like proteins is far more insensitive to changes in the components than the simple inorganic reactions we learn about in highschool. That is, the reactions depend mostly on the shape of the proteins after curling up, and on the amino acids on the "outside". And very often both of these will not be affected much by replacing amino acids "on the other side of the chain" by different, mutated ones. For example, they could be locked up inside once the protein has curled up. (q: Is this a fairly accurate general statement? I've read elsewhere that a small change in amino acid sequence CAN severely destroy the structure of the curled-up protein, but would this be in a minority of cases, or is it that despite this would be the case in the vast majority of cases, there's still an astronomical amount of possibilities to have a mutation NOT influence the structure?)
I don't know about astronomical but there is certainly a large potential for substitutions to be neutral. The most obvious issues for amino acids on the inner core is a change from hydrophobic to hydrophilic,which can radically disturb the proteins conformation, as in the case of the mutation in sickle cell anaemia.
I hope these help answer some of your questions.
TTFN,
WK
Edited by Wounded King, : Changed heading
™ Version 4.2