The origin of the two major theories of truth mentioned in this discussion (coherence, associated with constructivists, and correspondence, associated with instructivists) was with regard to the CERTAINTY of our knowledge (not with regard to probability of beliefs). There has been relatively little discussion within the instructional science community about the distinction between certain knowledge and probable belief--the interest seems to be almost entirely with regard to what could be considered belief systems. In that case, it is not a theory of TRUTH that is pertinent; rather, what are appropriate are evaluation criteria for belief systems, as perhaps elaborated by Quine and Ullian in Web of Belief. One pragmatic criterion for a belief system could be called its utility--does it help us do/understand things which were difficult/impossible before?
It may also be worthwhile to distinguish between ontology and epistemology, as well. Dave's ID2 seems to me to be founded on an ontology--there are in the world three basic kinds of things: entities, activities, and processes--that then determines what there is to learn about and how to design instruction to facilitate that learning. Unless I have misunderstood or misrepresented ID2, this view places ontology in the central and founding role with regard to the design of instruction. This starting point has some very discernible advantages: it has concrete implications and these implications can be and are being tested. Those with a genuine interest in our discipline and an openness of mind are likely to find value in the development and testing of Dave's theory.
Other instructional researchers (Van Marcke, for example) start from an epistemological foundation. In general, these researchers view learning as the acquisition of expertise and focus on various aspects of expertise and its acquisition. Van Marcke's particular interpretation also has very concrete and testable implications. As it happens, there is a great deal of agreement within these very different perspectives. I am not claiming that one point of departure is preferable to another. And I am quite certain that there are many other different beginning perspectives, as suggested by Tom. I would simply like to see more decent treatment of those who left from different departure points and have wandered through different deserts.
So, one could be an ontological solipsist ("only I exist") or an epistemological solipsist ("the only things I can KNOW are my own ideas"). These two extreme philosophical positions each suffer different problems: the ontological solipsist is not in a position to know the truth of "only I exist," and the epistemological solipsist has not said anything interesting (that could be refuted). Moreover, there appears to be very little utility for either view--not in any philosophical theory nor in an instructional theory.
There is philosophical extremism, as Dave indicates. There is also scientific extremism, as others suggest. What I find most extreme is to proceed from "I cannot imagine how..." to "it cannot possibly be the case that..." Would it not be strange for the universe to conform to the limits of our imaginations?