As Clark knows, I share most of his views and interests, so I'll be brief and focus on what appears to be the kernel of his thesis. I hope my concerns are not simply about the words used.
[quoting Quinn's paper] I contend that, at least for cognitive skills and at least at a coarse level, learning approaches are converging on a model... This model includes [a] motivating the learning by demonstrating the practical applications and importance of the knowledge...
Isn't motivation really something quite different from merely demonstrating the practical applications and importance of something? None of the research/theory on motivation, of which I am aware, suggests that motivation stems from demonstrations of importance, practicality, or applicability. Instead, motivation (that which determines what action an organism takes) seems to come from sources within an individual.
[b] providing a conceptual description of the skill...
Isn't this didactic instruction at its best (worst)? In what sense are conceptual descriptions engaging?
[c] demonstrating the application of the knowledge to practical problems...
How is this different from "...demonstrating the practical applications and importance..."?
[d] providing practice opportunities with support in the form of scaffolding...
Perhaps--if you are teaching a skill. But practice implies repetition that is both necessary for, and that leads to, mastery. But what about cognitive goals, for which practice may be irrelevant? How does this element contribute to a model that improves upon "drill and kill" (or does it)?
and, [e] facilitating transfer through guided reflection on the activity to integrate the practical issues with the underlying conception.
Guided reflection? Hmm. I like the sound of that. What is it? How does one guide reflection (a presumably wholly personal or subjective experience)? What is the relevance of reflection to practical applications? Is Clark seeking an operationalization of "understanding"? If so, how would it be measured? Of course, transfer would only be relevant if there is both a simulation of something and another something that is different from it or more "real"--e.g., in cases where the learning is not of the thing itself, but a representation of the thing.
My sense is that the model is contradictory in some elements and ambiguous in others. My sense, too, is that it is too broad, being, as Clark mentions, "spread across" problem-based learning, cognitive apprenticeship, pragmatism, etc. I'm not sure how much of these three areas are informed by research on motivation, but from my reading of the literature on motivation and cognitive apprenticeship, at least these two seem to be addressing different problems (cognitive apprenticeship is primarily a learning approach while motivation may be studied quite apart from learning). I would suggest that we need a model that is much more limited and specific if it is to help us make design plans and decisions, at least for instruction that is intrinsically motivating.