The thrust of a jet turbine is not produced by the jet efflux impinging on the air behind the exhaust nozzle. Thrust is purely a reaction to the acceleration of gas through the engine itself. Thus, all the magic has happened by the time the gas reaches the nozzle of the engine. The thrust vector originates at the engine itself.
Not quite...
The action/reaction in the "system" is between the airplane and the rest of the world. As is clear with the Harrier as well as any jet airplane with a controllable exhaust nozzle, changing the position of the nozzle grossly affects the thrust vector. It is the point of departure from the airplane -- i.e., the end of the nozzle -- that determines the thrust vector.
Note that the attitude control jets in the Harrier also use air from the "cold" section of the engine (the fan). If your "originates at the engine itself" theory were true, attitude jets fed from the same source would not function. However, they do, and the distance of one of those jet nozzles from the center of gravity of the airplane determines its lever arm.