Please allow me to ask a few questions of those with a good knowledge of biology (I'm just a radiochemist and not expert in these things).
I seem to remember seeing in a thread somewhere that some copying errors (mutations) result in duplication of genes; is that right? If it is right, I suppose that could "restock" the reduced diversity that Faith is concerned about as the copies diverge from each other with time; is that right too? As far as I can see, this reduction is Faith's entire argument.
Finally, I'd just like to note the "effect of large numbers" that seems to be missing in this discussion. If there are a modest number of changes happening constantly over a long time in a large number of organisms, that must amount to a very large number of trials. If only a few trials are viable, one would surely expect persistant changes from time to time. The need for the large numbers would make it hard to demonstrate in a short time, especially if any particular change is a matter of chance.
I am beginning to suspect that this discussion is the "are there beneficial mutations" discussion in new clothes.
Please feel free to expose any naivetes.
After all, comfort can be just about anything, provided you need it enough at the time. Grey Owl