Behe asks a lot of hard qujestions, sure. But, he also makes unfounded assumptions regarding the answers.
It’s important to note that Behe engages in a bit of mental slight-of-hand, to mix metaphors. He points to “irreducible complexity” as a challenge to Darwnian mechanisms, then uses specific pathways as his evidence. or, more correctly, a lack of knowledge of specific pathways.
There are two major components to what is generally referred to collectively as “evolution”: the mechanism, and the pathway. The mechanisms are processes such as natural selection, sexual selection, genetic drift, etc. The pathways involve the specific historical events which led to an end product, be it the vertebrate eye, a bacterial flagellum, or an entire species of ape.
What Behe does is to point out that there is an “unknown” pathway – how did flagella evolve, for example – and extrapolate from there that, since the evolutionary pathway is unknown (or, in his terms, perhaps, “unexplainable”), the mechanism must therefore be incorrect. This is the essence of “irreducible complexity”.
Unfortunately for him and his proponents, it is not logically sound. That we cannot describe a specific historical sequence does not in any way alter the fact that there are historical sequences for which we do have ample evidence. Thus, his argument becomes ad hoc – IC might apply in this case, but not that, and it’s anyone’s guess who gets to determine which is which. When a pathway is finally discovered, another target is found: “Yeah? Well what about…this, then?! Huh?!”
Behe also makes the mistake, whether through design or through ignorance, of neglecting as a possible explanation the fact that a complex structure can easily be assembled piecemeal if the intermediate structures performed different tasks than the final product. An example is the oft-mentioned bacterial flagellum, which, aside from the fact that it’s a poor choice for being irreducibly complex*, can be explained from the perspective that components may have once acted as secretory units. While it may be Behe’s criterion that structures must evolve piecemeal in a Darwinian fashion, with each intermediate form performing a “proto” version of its final function, it is clearly not nature’s criterion. Structures can evolve and change function as they do so.
*Behe delights in pointing out the complexity of one of the most-complex versions of flagellum, by citing 60-odd parts that all work together. He conveniently neglects that there are several much-simpler flagella which get by with significantly fewer parts. This would seem to imply that it is not, in fact, necessary for all 60 parts to be present, thus rendering the structure *not IC.