You are talking about natural mutations here, but more often, people talk about natural selection. What I am arguing is that genetic engineering is replacing natural mutations with artificial mutations (i.e. genetic engineering).
I am talking about a very specific type of artifical mutation. I am talking about changing the genome in a very specific way. For example, it is entirely possible to change one base and one base only in a genome. This contrasts with random artificial mutations caused by exposing organisms to x-rays, mutagens, or other means of random mutagenesis. Artificial merely indicates that man did it, not what the results are. I am more focused on the results.
Evolution is currently about natural mutations, which create new variations and natural selection which weeds out the bad while leaving the good.
More importantly, selection (both natural and artificial) allows neutral mutations to pass through at a probabilistic rate. Future mutations will interact with these neutral mutations, resulting in phenotypes that may not have been predicted through a rational design approach towards genetic engineering.
With genetic engineering, humans will create organisms, but nature will still stamp the organism as fit or unfit.
What if we base fitness on genotype and not phenotype? What then?