As I understand it, you are begging the question since WAR isn't a measure of a team's record with a particular player in the lineup, but how many wins he will contribute to the bottom line of the team by being in the lineup. Trying to extrapolate from the team's record his contribution seems about like catcher's ERA to me, whereby you are trying to make a correlation where there is not one.
The argument might be better stated that adding a 5 WAR player to a good team might have an exponential effect on the team's record. Add that 5 WAR player to a below average team, and they will improve, but not as dramatically. And I state that as a hypothesis.