Ah, I think I misunderstood your previous post. You're basically asking why a designer didn't toss fish genes into a mammal, to make a swimmer.
The obvious counter question: if a designer wanted to make a swimming mammal, why would he start with a fish, instead of a mammal?
Note, BTW, that there is no reason for us to rule out the possibility that natural selection is operational at the same time as, and in between instances of, intelligent intervention.
Which is exactly my point; if they do the same things, why aren't they made of the same stuff?
There's no requirement for them to be made of the same stuff. All this objection does is tie the hands of our putative designer, thereby arbitrarily disallowing certain areas of the design space.
The differences between fish and whales are typically 5 - 10 times as large as the difference between hoofed animals and whales.
Again, this not an argument against design. Indeed, if for the sake of argument we granted the existence of a designer, we would not be at all surprised to see him working gradually from some fixed starting point. After all, that's exactly how selective breeding works.
A homely example: if you wanted a Big Dog, would you create same by splicing horse genes into his DNA? Or if you wanted a Small Dog, would you splice in mouse genes? Although you probably could do it that way, you probably wouldn't, because you also want the result to remain a dog.
Direct genetic modification works that way, too. Scientists desire corn with some new trait, and they stick in a slice of genetic material that confers the trait. The basic genetic structure is still far more similar to corn than it is to whatever donated the genetic structure.
As you've pointed out, these days the inserted genetic material is obviously non-corn -- but that is primarily a function of the current state of technology: it's easier to borrow than to build from scratch. In future, though, is it not likely that a more advanced geneticist would want to massage the corn genome directly, and thereby bypass the difficulties associated with inserting foreign genetic material? (Your own posts suggest the distinct possibility that this will in fact occur.)
What all of this has left out, of course, is that with design comes "reasons to design." We know that with humans, intelligent intervention is often motivated by some practical result, such as more grain per wheat stalk, or miniature horses for use in coal mines. Other times, as with teacup poodles, the motivation is basically impractical. (As the psalmist put it of leviathan: "which you formed to play in [the sea]." Psalm 104.)
Because fish, like marine mammals, are adapted to life in the water. Physiological adaptations that enable them to live in the water are not needed by land animals. For example, whales have adaptations that allow them to go without breathing for far longer periods than land animals are capable of, but not for infinite periods. Why not give them gills?
A homely example: if you wanted a Big Dog, would you create same by splicing horse genes into his DNA? Or if you wanted a Small Dog, would you splice in mouse genes? Although you probably could do it that way, you probably wouldn't, because you also want the result to remain a dog. Direct genetic modification works that way, too. Scientists desire corn with some new trait, and they stick in a slice of genetic material that confers the trait. The basic genetic structure is still far more similar to corn than it is to whatever donated the genetic structure.
Actually, most of the time that's exactly what you do. When you want a roundup-resistant corn variety, you splice in a gene from a soil bacterium. A single gene does not make the corn into a bacterium. Much of your argument is that we already know that intelligent design is used - by humans - to modify organisms: why assume it never happened before we came along? Well, the answer is, we don't see the idetifiable foreign genes in native species, that we see in genetically engineered crops.
This link has a decent synposis of how we make transgenic corn.
This link has a list of transgenic organisms. Note that we've put pig and human genes into fish, cow genes into bacteria, and so on. We don't put dog genes into dogs, because generally the dog wouldn't change much.
In future, though, is it not likely that a more advanced geneticist would want to massage the corn genome directly, and thereby bypass the difficulties associated with inserting foreign genetic material?
DNA is DNA. If you want to insert a working enzyme, you'd find one in nature that you know works, and slice out the thousands or so base pairs needed to code it. It's 'foreign' in that it came from another organism, but the host organism won't know it's foreign. You wouldn't try to invent one de novo. In fact, even new enzymes in nature are found in general to have been modifications of pre-existing enzymes, not new creations.