Behe writes here about the fuzzy hypothesizing (don't worry--it's tolerated because it doesn't threaten neo-Darwinism) on the complex interactions within cells--particularly catalytic networks.
Irremedial complexity tries to fuse the non-selective aspects of Kimura's neutral mutations with the "information for free" notinos of Kauffman's hypothesis of random catalytic networks. Kauffman is hoping to fill as large a gap as (or much more than) Kimura's idea did. As Behe notes, "irremedial complexity" is a possible solution to a possible problem (i.e. it will only be acknowledged that there was a problem if it turns out to be a solution) that provides a narrative for the seemingly gratuitious complexity we observe in cells--especially eukaryotic cells.
Seemingly gratuitous? Like junk DNA was seemingly useless? Sounds like a vestiges-of-the-gaps argument to me.
Irremedial complexity: Autologous name? Is this an irremedial concept? Who knows? Maybe it is an idea worth considering. But considering the vitriol against ID for not coming up with exact figures for everything with exact justifications (is anything really that rigorous in evolutionary biology? the figures are all over the place...) that every biologist accepts, isn't it curious that "the literature" is tolerant of possible solutions to possible problems?
It is funny that they describe bureaucracy as an ratcheting, irreversible complication of control. But that analogy isn't encouraging that anything good would come out of it.