It's a tricky situation. I agree with most of this article. Examining these complex structures and trying to understand how they could have come to be using mechanisms that we've seen evidence for (like separate structures with totally different functions coming together to do yet a third thing) is valuable. Hence, if some of Behe's results were actually verified, that simple genetic mutations can't account for certain structures, that's valuable, because it shows that these complex combinations of structures must occur at a useful level in natural selection, which is, well, pretty cool, in my opinion. It's sad that people keep trying to "use" these results in various ways to advance particular agendas.
That's kind of what I took away from this article, but then I've been guilty of reading what I want to read sometimes...
Yes, very valuable. But how do you show that? So far, Behe's examples have all been failures:
Behe's "irreducible complexity" argument is fatally flawed. Ichneumon's post 35.
Irreducible Complexity Demystified. Major debunking of ID.
The Flagellum Unspun: The Collapse of "Irreducible Complexity," Kenneth R. Miller. Critique of Behe.
However, if there were something like, for example, an animal with one limb that had a functional M-16 built into it, and there were absolutely nothing remotely related to that in the fossil record, then we'd have to agree that it would certainly have the appearance of design. But nothing comparable is ever discovered.