ID has consistently, however, run into one specific roadblock: the concept of methodological naturalism. Since ID requires a supernatural component at some level/point, it is difficult to say that it is truly a scientific pursuit and retain any credibility (not that Johnson has a surfeit in that department).
So folks like William Dembsky, Sal Cordova, and the other Disco Institute buffoons tried their hand at arguing that methodological naturalism is flawed or unnecessary. Still, no deal.
So, Johnson has gone back to his roots and is inventing new terminology to dress his tattered old ideas up in: this time, he takes on naturalism using "Theistic Realism", which is really just "goddidit" gussied up a bit. Allen McNeill over at The Evolution List has an excellent analysis of this argument, and why it fails. It's worth reading.