Though much has been made in Creationist circles of the supposed 'irreducible complexity' of certain fundamental cellular structures, this argument is not really of concern at all to the greater question of evolution by natural selection. For almost all of evolution can be boiled down to simple changes in the length and thickness of bones, and the consequent size and placement of a common set of muscles and organs supported by these bones. And a great proportion of the mutations which account for evolutionary change are simply copying errors wherein the usual course of DNA replication results in the forces of transcription generating of in extra copy of a link in the chain -- exactly the sort of change which best accounts for variations in bone length. And consider this, there's no chemical difference whatsoever between, say, cat and dog DNA, or fish and human DNA (or even still between the DNA of a fungus or a ficus and your friend from Finland). Now in every case, DNA is composed of strings of four specific chemicals, coming in a certain number of potential pairs. Perhaps that is an inherent factor of our Universe, the fortuitous ability of complex molecules to form at all, and one which suggests a designer.

But there is no need to extrapolate from that possibility a wholesale creation of living things in their currently existing distributions. Consider this, if you bred dogs, and every generation you picked the most dog which was most catlike in terms of the length and thickness of its bones (and perhaps a few other visual cues, ear shape and such), in a few thousand years, you could breed a dog into something which looked exactly like a cat, and with only a bit more selectiveness along the way, you could breed one which looked and acted like a cat. And, more amazing still, in a long enough time, perhaps a million years, you could actually breed a dog into a cat at the genetic level. But before needing to get to identity of DNA, the initial and apparent difference between a dog and a cat is length and thickness of bones.

And remember that it only took a few thousand years to breed borderline feral wolflike canines into a world of shihtzus and chihuahuas. If men bred chimps and selected for human-likeness, we could breed chimps into humans (a second time) in a million years as well. New humans able to reproduce with the old ones, for if you selected for similarity of DNA, eventually the DNA of the breeding population would indeed be identical to human DNA. But, naturally, though evolution sometimes works towards convergence in appearance, it never works towards convergent DNA -- that has never been observed, even though it would be a true creationist miracle. In nature, once speciation has occurred and populations cease to be able to interbreed, they can never go back, so absent intentional engineering, humans and chimps (or, wolves and foxes, or emus and ostriches), will never evolve naturally to interbreed again.

But intelligent design could indeed bring about convergence in the DNA of any two things which have DNA. All it takes is time. Give me a billion years. I will breed a sea cucumber into a wombat. And back again!! And for the most part, I'll be simply choosing for the length and thickness of bones. Once I get it to grow bones, that is.