Basically, the hypothesis proposed by Pallen & Matzke, is that the bacterial flagellum arose stepwise from a simpler, antecedent system, possibly a system that had nothing to do with motility in the first place.
The papers in question destroy Behe’s assertions via a two pronged attack. First, by demonstrating that a flagellum can still function while missing several components. A particularly juicy find in one of those papers, centres upon the FliI and FliH proteins. Knock out the gene for FliI, and the whole process of flagellar biosynthesis crashes to a halt. But if you THEN knock out the gene for FliH, flagellar biosynthesis restarts, and the resulting flagellum works just as well as the original.
Next, the papers demonstrate that the genes for the various flagellar proteins aren’t some unique and ineffable construction arising from nowhere (oh, the irony of creationists asserting this …). Instead, the papers demonstrate that numerous homologies exist with other genes, indicating that the entire flagellar system arose via co-opting genes from another system, then adding some new features to result in a system with a new function.
And, lo and behold, it has been found subsequently, that a large part of the bacterial flagellum’s genes and proteins, were co-opted from a totally different system. Namely, the Type 3 Secretory System (T3SS). This is the system that allows some nasty pathogenic bacteria to inject lethal toxins into your cells.
The T3SS constructs a sort of hypodermic needle on the surface of pathogenic bacteria, and is found in such organisms as the bacteria responsible for typhoid fever, bubonic plague and cholera. Since the T3SS already contains features that would be readily reusable for the bacterial flagellum, biologists have been working on what evolutionary pathways would facilitate the relevant transition.
At some point, I’ll have to do a literature search, and see if that investigation has borne the requisite interesting fruit.