I'm not sure these are totally incompatible. In current thinking, it seems the belief is that:

Markets create an efficient APPROXIMATION of perfect resource allocation, and groups of humans can also APPROXIMATE resource allocations too, but with weaker precision than a whole market, i.e., a larger group of humans.

So the two lines of research can be made to fit inside the description I just gave. One comments on the glass-half-full portion of market approximation, and the other, on the glass-half-empty portion.