I think, however, that both Behe and his opponents are on the wrong track. They fail to distinguish the need for causes which physically produce the object from causes responsible for the plan. Planning and actually constructing may be found in one agent, but need not be. The architect may never touch the house, but is certainly responsible for its construction; whereas the artisans who actually assemble the house may have no idea of its overall layout, but simply follow instructions. One might legitimately ask: Are those who deny design doing so simply because they have found agents which account for the house's completed assembly? Are those who are trying to defend design, seeking to show that there are no natural causes capable of assembling the effect so that a Designer has to directly intervene? It is not hard to see that it is one thing to have the 'smarts' to plan something, and it is another to have the physical power to realize a plan. For example, insects clearly have the power to produce ordered results, but they lack intelligence.
To put it another way, Behe (and, of course, he is not alone) muddles together different sorts of causes -- in particular, generating causes and exemplar causes. There is certainly an argument to be made that generating causes presuppose some sort of exemplary cause, the latter as it were etching out the possibility of the things in question coming to exist at all. But the question of whether we can have a Darwinian explanation of a type of biological system is entirely irrelevant to this argument, since at that level of explanation we aren't considering exemplar causes, just (proximate) generating causes (in particular, efficient causes that explain how a given type of organism arises and is perpetuated). And this is why there is a dangerous ambiguity in much ID talk. Things may 'exhibit design', and, I think, certainly do. But to exhibit design is simply to be related to an exemplar cause. What IDers are really advocating is that the exemplar cause is also the (proximate) generating cause, and we have no particular reason to grant that. It isn't impossible that it could turn out to be so -- as some IDers have said, it could very well be that we will turn out to find that intelligent aliens (for instance) have directly genetically engineered life on earth. That's possible, but none of our evidence points in this direction. The relation to an exemplar cause is important. It has a role to play in metaphysics, for instance, when we ask the question of why things are possible at all; and it would be the ultimate foundation why (e.g.) thinking of biology in engineering terms can be so useful. But it is not, at least as far as our evidence indicates, actually relevant to the sort of questions IDers want it to answer. They only think it is because they are equivocating. This equivocation is why the occasional appeal on the part of IDers to the fact that biologists use engineering-type reasoning, and treat organisms as designed (or as if they were designed), doesn't actually strengthen their case. It just shows that it is useful to think of effects in light of an exemplar cause (whether one does so in the belief that the exemplar cause actually exists, a claim for which I think good arguments exist, or just as a heuristic device, as most naturalists would want to claim). The real core of ID, however, is a very controvertible thesis about the relation between an effect's exemplar cause and its proximate efficient cause. And that's a different sort of thing altogether.
George also makes an interesting comparison between ID and occasionalism, although I don't think she develops it quite correctly. Occasionalism is the position that there is only one true cause (God) and all other 'causes' are really just occasions for the action of the true cause. ID is not occasionalism, but considering ID in light of occasionalism genuinely is interesting. The purpose of occasionalism was to give greater glory to God; by stripping creatures of causal power, occasionalists thought they were giving God His due. However, while you can be an orthodox Christian and an occasionalist, orthodox Christians almost never are. Occasionalism fails because it makes the mistake of treating primary causation (that of God) and secondary causation (the sort treated in, say, physics) as contraries. But they are not. Aquinas gives the classic statement of the problem by noting that occasionalism detracts from divine omnipotence in holding that it could not communicate a real causal power to creatures. Short of actually demonstrating that the attribution of causal powers to creatures is a contradiction, the occasionalist can't really claim to have given greater glory to God than the person who thinks creatures have real causal powers, for the latter denies no power to God that the occasionalist affirms, but the occasionalist denies a power to God that the latter affirms, namely, the power of giving power to creatures.
It is worth noting that the same criticism of occasionalism is found in Hume, who in ECHU 7.1 makes a remark with which Aquinas would fully agree:
Thus, according to these philosophers, every thing is full of God. Not content with the principle, that nothing exists but by his will, that nothing possesses any power but by his concession: they rob nature, and all created beings, of every power, in order to render their dependence on the Deity still more sensible and immediate. They consider not that, by this theory, they diminish, instead of magnifying, the grandeur of those attributes, which they affect so much to celebrate. It argues surely more power in the Deity to delegate a certain degree of power to inferior creatures than to produce every thing by his own immediate volition.
For Aquinas this is a substantive criticism, whereas for Hume it is a criticism of the internal plausibility of occasionalism (i.e., one of the things that makes occasionalism plausible to many people is something occasionalism actually doesn't have). But the basic point is the same. This is one of the reasons why occasionalism is not a popular theory of providence among Christians. The same reason that makes occasionalism unpalatable, however, should also make Christians wary of the sort of reasoning put forward by ID theorists. It's no secret that, while ID doesn't strictly imply Christian theism, one of the reasons why so many Christians are sympathetic to it is that it seems to them to cohere well with Christian theism. However, in that light the same problem with occasionalism arises with ID reasoning: short of actually demonstrating that it is a contradiction for creatures to have the particular causality under examination, ID doesn't actually yield what Christians have often thought it yields (it forces us into claiming a weaker, not a stronger, role for divine power in biological processes). The worry is weaker against ID than it is against occasionalism (chiefly because occasionalism is a completely general position, whereas ID is focused on a particular sort of case), but it should be there. The problem doesn't affect ID as such, but it should at least raise questions for many of the Christian supporters of the movement, questions that need to be considered but are not usually discussed.
There are other problems with ID reasoning (e.g., problematic presuppositions about chance causation), but I wanted to bring these two up because they are interesting, serious, and rarely considered.