PT readers may be interested to check out this great new article in New Scientist, which reviews recent developments in flagellum evolution. The thing I find interesting about all this is how the IDists have been intellectually unable to concede any tiny little mistake in anything they said, e.g. this standard ID argument from DI fellow Bruce Gordon as it was presented in 2006:
As the biochemist Michael Behe has pointed out, this flagellar motor depends on the coordinated function of 30 protein parts and it will not work if even one of them is removed – it is, in his terminology, “irreducibly complex.” Since natural selection works (in neo-Darwinian theory) by environmental “selection” of functional advantages manifested in the phenotype that have arisen through random genetic mutations, it can select the motor once it has arisen as a functional whole, but it cannot produce it in the step-by-step fashion required by neo-Darwinism because every stage of lesser complexity is completely nonfunctional.
Now, the best you will get from the ID guys are evasions like “oh, we never claimed that subsystems would be nonfunctional.” See above, guys.