I'd like to step down from the moderator's chair for a bit and say a few words about Brent Wilson's paper.
First, I think Brent is one of the best writers and most original thinkers in our field. We are lucky to have him in our profession and on this Forum. I enjoyed reading this paper very much and thought many of the ideas were very interesting.
Second, I have struggled with many of the ideas Brent puts forth in the paper. I found the paper somewhat analogous to Tom Reeves earlier ITForum paper "Questioning the Questions of IT Research" which is another must read, I think, for anyone in our field.
But the real question raised by this paper, and I think it goes unstated , is not whether theory-based design is bad, but whether or not we have adequate theories on which to base our designs.
It seems to me that, as a field, we had a great deal of success with some of our early attempts at theory building. Brent mentions the work of Gagné, Merrill, Glaser, Scandura, Briggs, Tennyson, and Reigeluth. I think the problem is that we have done a terrible job building on those theories. We seem to have accepted the work of those researchers as gospel and broken up into camps to debate which is the true gospel. We haven't taken that next step--pushed on to explore the frontiers opened up by earlier research.
So, in a nutshell, I don't think the problem is theory, I think the problem is that we haven't done enough research to build a suitable theory, yet. I guess there is some kind of continuum in any field that runs from those who think a unified theory is possible and desirable, to the other extreme in which people think a unified theory is impossible and undesirable.