I don't agree with so much in your comment, that I have to reply, sorry. First of all, we don't start with the same gene set. Ok, maybe we did 4 billions years ago, but it's a pure speculation and we don't have any instruments to infer anything valuable from the fact that 'there was a single organism at some point of time from which everything evolved'. I'm pretty sure the whole picture is much-much more complicated. As for humans, I can say with a high confidence that there was not a single human being evolved that become the genesis of all humans. Specification is a complex thing and there's always a period of hybridisations with closely related species. I won't delve into it deeper, but some fishes even rely on other species for their reproduction. This fish is actually really amazing and there's a lot to unpack with its reproduction. Secondly, there's no such thing as random distribution that does something for organism fitness. The species that got an appropriate set of genes (and maybe even more than genes, like some epigenetic markers) survived. The ones who didn't, didn't. That's it. In many organisms there's not a unique set of genes ("formula") that lets it survive in a given environment. Gene regulation is incredibly intricate and has a ton of feedback loops. For example some genes, that are very important are often duplicated, like ribosomal or histone genes, and mutations in them doesn't do much.

Both coding and non-coding DNA. Actually, 0.1% is a little bit outdated. The variance can be higher according to the 1000 genome project. It is said in this article that > We find that a typical genome differs from the reference human genome at 4.1 million to 5.0 million sites. Although >99.9% of variants consist of SNPs and short indels, structural variants affect more bases: the typical genome contains an estimated 2,100 to 2,500 structural variants (∼1,000 large deletions, ∼160 copy-number variants, ∼915 Alu insertions, ∼128 L1 insertions, ∼51 SVA insertions, ∼4 NUMTs, and ∼10 inversions), affecting ∼20 million bases of sequence.

These ~20 million bases count for ~0.6% of total genome length