Uhm that’s an extremely misleading citation by Behe. The article is estimating the rate of fixation of gene-duplications, not their rate of de novo occurrence. Those are two very, very different things.
As they say:
(Gao and Innan 2004): Note that by “gene duplication rate” we mean the rate at which a duplicated gene is created by mutation and becomes fixed in the population. The fixation probability of duplicated genes should be largely affected by natural selection (14).
Behe is trying to argue that there is some sort of problem with the timescales at which we can expect proteins coding genes to duplicate and diverge. But to do that he needs the rate of de novo occurence of gene duplications, that is the rate of mutation. Not the rate at which such duplications ended up becoming fixed in the population by natural selection and genetic drift. The measure produced in that paper is actually more accurately described as an average rate at which duplications have been beneficial in yeast.
In any case there is no such thing as the rate of occurrence of gene duplication. It is highly variable depending on locus. In an offshoot experiment of the Long-Term-Evolution-Experiment they propagated the Cit+ mutant in a citrate-only environment (no glucose). In this environment gene duplications of the Cit+ locus and numerous other genes involved in citrate metabolism were extremely beneficial, so much so that within 2500 generations some strains had increased their total genome size by over 20%. In one case a particular gene (maeA) was duplicated over 95 times. Look at table2 in this paper:
All false. Completely false.