First as a rule: effective thrust is created at the exit point of the exhaust gases. If a turbine pushes 50lbs, and looses 10lbs through a pipe (exagerated), the turbine in front will most likely still be producing around 50lbs. The 10lbs can be found along the path of the pipe, pulling the other way.... (trying to paint a very very rough picture: the turbine pulls forward with 50lbs and the pipe pulls aft 10lbs, thus pushing the airframe 40lbs in total).

The theory behind it is not so very much difficult if abstracted a bit: take a long straw and blow into it. Air will exit at the other end very quickly. Now take a large diameter pipe and blow into it. Exhaust velocity at the other end will be miserable.
Energy, and thereby thrust, is defined with mass x velocity². You see, by lowering velocity you severely hurt your energy. By increasing mass you slightly increase thrust.
When the air exits through the nozzle and finds no pipe to keep up pressure and speed, it will expand more and more. Air surrounding the nozzle will be ripped along also (venturi effect), slightly increasing the mass, and at the same time slowing down the larger the pipe gets.
Most air is ripped along right at the nozzle, the further out you go the less the effect....so the larger you get, the less your gain in mass is. At the end of the airplane, where the effective thrust is created (total amount of air x velocity²), there will be less remaining to push ahead.
In an open air situation, thrust is also created right where the airplane ends: in this case right at the turbine nozzle. So you get exactly what the turbine is producing without losses.

The reason why the wren pipe "increases" thrust is also very simple: to a certain point venturi air cause little loss in speed but relatively high gains in mass flow (remember the part about most air being ripped along right at the nozzle?). The losses in velocity are more than accounted for in the gain in mass within this amount of expansion.
Slightly expanding the pipe as Wren does allows just this to happen in a controlled environment, thus gaining STATIC thrust. Dynamic thrust (moving airframe) will decrease though.....so basically you are buying more static thrust for less top-end performance. Since the exhaust velocities are so high, most modeling uses will not notice this effect at the models top speed.

Hope to have shed some light on this mystery...
Best regards
Hank