Here Johnson goes on to attack the mechanism Darwin proposed for the process of evolution. I think it's safe to say that there's little reason to think he does a better job of understanding the facts behind the theory.
First Johnson starts with another error stating that variation (spoken of by Darwin) is the same thing as mutation (p. 16). He admits that he is lumping a number of different processes under this term. It turns out that many of the processes Johnson calls mutation are definitely not. Recombination is a process involved with genetic re-shuffling due to sexual reproduction. It is not the same thing as mutation, which is the production of novel genes by random processes. While recombination can produce new combinations of genes, it certainly isn't the same thing as mutation. This is a basic fact of biology, and not understanding this concept doesn't say a great deal for Johnson's understanding.
He notes that a great deal of the evidence Darwin presented was the variation produced in domesticated animals under human breeding patterns. He notes that such artificial selection is based on intelligent beings making the selection and produces more fit organisms. However, this is really a misunderstanding. Just because an intelligent being is doing the selecting doesn't mean the selection is any different. Natural selection says that any organism that is a better match to its environment will survive better. Last time I checked, humans are part of the environment, and any pressure they put on any animal to change its breeding is just a part of the environment. For example, human use of antibiotic has put a selective pressure on bacteria to produce strains that can resist the medicine. Now, how does that differ from a person trying to breed a more resistant strain of plants? In the latter case, the human has a goal in mind, but that's not relevant. We see that even when there is no goal in mind, humans are able to produce evolution in other species. It seems that there is plenty of evidence on evolution occurring both in "artificial" and "natural" settings.
Johnson goes on to repeat the creationist claim that laboratory experiments have produced new species of fruit flies, but that they remain fruit flies (p18-19). He notes that dogs don't become as big as elephants because they lack the genetic variation to grow so large. He says that while nature has millions of years to produce new forms, the extreme selection produced by humans in labs should be equivalent. Nice try, but selection requires variation to act. For evolution to occur, there has to be genetic variation. Problem is, in the few thousands of years humans have been breeding dogs there will have been very few mutations arising that weren't already present. It doesn't matter how strong the selective pressure is, if there isn't some variation present in the organisms, then no evolution can occur! To claim that human's performing selection could somehow magically produce new variation is not only ridiculous, it's idiotic! Either Mr. Johnson has even less of an understanding of biology than I thought he did or he is being purposely deceptive. To claim that he can evaluate the logic of evolutionary arguments without even understanding the most basic biological facts is silly.
Next he pulls out the old "natural selection is a tautology" argument. He defines natural selection as "survival of the fittest" (which many people have done before him) and says that since the fittest are those that survive, the statement evaluates as "survival of those who survive". This is a worthless tautology, and rightfully so it would not be a useful definition. He says that it's impossible to identify whether a particular trait is an advantage outside of the increased survival it confers. This is not the case at all. It's quite possible (for at least some traits) to know enough about the environment and the organism to say which of several traits is advantageous before hand. If we observe 5 individuals, each with a different trait, we can make hypotheses about which trait is better in the environment. We do not have to define fitness in terms of survival, we simply use survival as a way of measuring fitness.
He then goes on to try and discuss natural selection as a scientific hypotheses. He cites a variety of examples that have been posited to show evolution in action. Examples such as industrial melanism in moths in Britain and changes in the beak size of Darwin's finches in the Galapagos. In both of these cases, the change that occurred was reversed some time after the intial change. This is true, but once again, Johnson is ignoring the definition of evolution. Evolution = change in genotype frequencies. This means that when the moths switched from being light to dark, that was evolution. When they changed from being dark to light, that was also evolution. To claim that this example doesn't show evolution simply points out that Johnson still doesn't know what the heck he's talking about.
Then he goes on to claim that natural selection is a philosophy with no scientific evidence to support it. Ignoring the fact that Johnson doesn't know what he's talking about with regards to the evidence, we should deal with this claim. He says that the lack of evolution in "living fossils" is a problem for evolutionists. I don't know about every one of these fossils, but I do know that at least in some of the cases the creature that is supposedly the same as it was millions of years ago is actually quite different. Check talk.origins for more information on this.
Next he complains that the theory of evolution is too malleable. He is actually upset that "if we assume that Darwinism is basically true then it is perfectly reasonable to adjust the theory as necessary to make it conform to the facts." (p. 30). Of course, this is what science does, adjust the theory to fit the facts, not vice versa! It would be bad, as he says to assume that Darwinism, but it certainly has evidence to support it, regardless of his claims to the contrary.
He ends the chapter talking about the peacock's feathers. He isn't able to understand why a peahen would choose a mate with a life-threatening tail. He says that he can't see how selection could produce such a trait, but that it seems obvious to him that a "whimsical creator" might favor just such a silly tail. Once again, Johnson is ignoring a great deal of scientific theory and debate. There are theories about how this occurs based on the fact that a male that is able to survive with such a handicap is probably a good individual. This means that a female can do well to choose a male who is able to survive with such a handicap. In Johnson's favor, this is currently a topic that has received a great deal of attention, and his book was published in 1993. However, this topic hasn't only been studied since then, so he is still guilty (again) of ignoring science working on the problem he perceives.
In summary, he's once again showing a failure to grasp the basic information in favor of attacking "logic" when he doesn't understand the underlying basis for the arguments being put forward.