I actually went and read a couple of Dembski's papers and articles on intelligent design at the link you posted. The problem with Dembski is that he doesn't actually prove anything. His primary tool is "proof" by analogy or anecdote, which is absolutely worthless with respect to supporting his position. There are also a number of subtle non-sequiturs in his arguments. I did not find a single real information theoretic argument; I only found vaguely related assertions that appear to be made from an invalid premise (though I clearly get the impression that he doesn't actually understand information theory).
Overall, I was underwhelmed. Is there anyone else who does a better job of constructing an intelligent design argument than Dembski? Dembski talks an okay story, but his papers are full of obvious holes to my keen eyes, too numerous to itemize really. It reads like it was written by a salesman rather than a scientist, a distinction that would be lost on people not used to critically evaluating these things. If my clients asked me to evaluate that paper as a proposal of a new concept, I would give it a "no go". (I used to work as a consultant for VCs evaluating new technology proposals that were too steeped in scientific or pseudo-scientific banter for them to get a good feel for the legitimacy of it, a fun job since I got to put people through the intellectual wringer to separate the charlatans and kooks from the bright guys with original ideas.)