The way the author phrased this is unfortunate. However, there are very good reasons to believe that even small mutations could have resulted in large changes in morphology or behavior. A mutation in even a single base-pair could result in cascading effects. See: http://en.wikipedia.org/wiki/Homeobox
Instincts are coded for by genes. Genes are made up of DNA strands which use a sort of language, Gattaca. In English words, a man’s instinct may have been quite simple, and something like the following:
"Look at women, and see if they have a pair of round wobbly things on their chests. If they have such a thing, feel sexually repulsed. Increase this feeling as they are bigger, rounder, wobblier."
One might argue that a man with such an instinct would do well, in the light of the changes in the bodies of women, to delete this programming entirely. However, a much smaller mutation could alter the programming to:
"Look at women, and see if they have a pair of round wobbly things on their chests. If they have such a thing, feel sexually stimulated. Increase this feeling as they are bigger, rounder, wobblier."
Can we say "pseudoscience mumbo-jumbo" here? Just because it's a smaller change in a natural-language description of an instinct, it doesn't necessarily mean that the mutation is smaller. The author is making a huge, unwarranted and rather silly assumption about the relationship between the DNA structure and the complexity of the natural-language description of the consequences of that DNA structure.