I had a read through your article to figure out where your numbers came from, and there is all sorts of wrong going on there. Let's start with one wrongness that's simple to explain.
Since you enjoy logical fallacies, your calculation is colloquially known as the sharpshooter fallacy. You're arguing that since it's extremely unlikely for one gene to evolve; it's clearly impossible for many to evolve. But what you're calculating (wrongly) is the probability of one specific gene evolving. How likely that is is of course a very different question to whether any, non-specified functional gene could evolve.
Another obvious wrong with respect to your numbers is what you call the 'deformation tolerance' of a gene; meaning how many ways could you change it and still have a functional gene (where what you mean is the same function - see above). You use the assumption of 50% and claim that's way too big, and thus a generous assumption in favour of evolution. You present no evidence or reasoning for why that's too big, and it does not necessarily appear to be so, as discussed in one of the articles you cite a few paragraphs later.
In discussing DNA methyltransferases; it mentions that you can change most of the amino acids in these proteins and still end up with a protein performing the same function. So the 'deformation tolerance' is more than 50%. Add to this that we're talking about changing the amino acids in the gene product - many of the possible changes in the base-pair sequence of the gene coding for this product will not change the amino acid sequence. Your 50% assumption is not too high - it's woefully too small.
So, all one needs to do in order to refute my argument is to address two things: the deformation tolerance and unit of selection. If someone disagrees with 50 percent deformation tolerance he must explain why he disagrees.