I read the 2008 article as saying that a "map" of human psychology that does not including meaningful reference to internal mental states will be accurate enough to be useful only by coincidence.

Behaviorism does not include meaningful reference to internal mental states, yet the second article says that it can be a useful map.