The central theme of the Vehicular Cycling ideology seems to be J. Forester's pithy formulation: "Cyclists fare best when they act and are treated as drivers of vehicles". We have a traffic system, the argument goes, designed for a given set of behaviors. Bicycling on the road works best when the cyclist fits into that system with minimum fuss.
The problem I see with using this idea as a prescription for cyclists' behavior is that cyclists are often not treated as "drivers of vehicles". For example, other drivers often fail to yield right of way, or they yield it inappropriately. Also, some parts of the traffic system are not safely usable for bicycles--unresponsive signal triggers is an obvious example. Many other examples have been discussed in this forum.
My point is not to catalog all the ways bicycles have trouble with the existing traffic system. My interest is in the question of whether bicyclists should aim to behave "as drivers of vehicles" when they are not treated as such. I agree that the best outcome occurs when both predicates of Forester's statement are true, but it seems to me that when the cyclists are not treated as drivers of vehicles then they are wise not to act like them, either.