Well, I've lurked long enough with regard to this issue [converging models], and would like to get something straight with regard to Gagne, Briggs, and Wager concerning constructivism. Gagne, first of all, would tell you to your face that it is BS. Briggs, of course is dead, but as I knew him he was a futurist, concerned with affective and social outcomes of learning (both formal and informal). I think I fall mostly into the pragmatist camp. Well, that being said, Gagne was trying to define conditions that would lead to learning. With regard to these conditions, it is his premise that different kinds of learning require different instructional environments. He is big on "prerequisites" for learning, which might be equated to the idea of scaffolding and the readiness of the learner for learning a new concept. He also emphasized strategies appropriate for the type of learning outcomes you were designing for. If the goal was that the student would take personal responsibility for his or her own learning, well, that is an affective outcome, and the learning environment would have to be structured to support that outcome (it might even look like a constructivist environment). Furthermore, Gagne believes in generative learning (he called it higher-order rule or problem solving. Problem solving leads to the learner's construction of new rules, which can serve in future situations if the cues call it back into working memory. Never did Gagne, Briggs, or Wager say that there was one path to learning anything. The learning hierarchies and/or curriculum maps simply represent a knowledge structure that should be supportive of learning. One doesn't have to do many maps to see that there are many ways to structure complex knowledge to support learning.
The point is that we as learners use many different models for learning when they suit our needs. I, for instance, when learning a new computer language, depend on it operating something like other languages I use. So, I construct possible combinations, and structures that from my experience ought to work. However, if they don't, I quickly look at the manuals to "acquire" what knowledge I need, then I apply it to verify my understanding. I don't even need to work in a group to do this. Direct instruction does have its benefits, and eight tenths of the time, the companies I consult for are asking for better direct instruction or performance support tools to eliminate the need for instruction. Maybe if some of you read Gagne with an open mind, you would see that it very much supports generative learning, if that's what your goals are. Now, there is two cents from someone who believes in the traditional model of instructional design and constructivism concurrently.