True when both screws are turning, but the original question was regarding the causes of yaw when one engine fails (besides the obvious asymmetrical thrust).
You have a point there.
Basically the effects of torque act in opposite directions on the fuselage, one acts upwards, one downwards. They compensate eachother during normal operation on the ground, but indeed the engine failure case is different as the engines become the center of a left roll axis.
I got messed up about that.
So let' s discuss the engine/prop' s torque' s effects in determining a critical engine on a twin, on the ground.
In a few words: different explanation, same conclusion being little to no factor for critical engine.
If you' re not interested in the analysis, jump to the last paragraph.
Take an imaginary heavy twin engined turboprop with a very long wingspan and engines very close to the root. Due to the central positions of the engine, you can assume that it will behave like a single-engine airplane if an engine fails, including the torque creating a left roll moment, adding weight on the left side of the airplane and to the main gears on that side, increasing drag on that gear and yawing the airplane to the left.
Now take the same turboprop and place the engines at the tips of the wings. The engine/propeller combination produces the same torque for a given configuration and specific air data.
Torque is T = r x F, r being your arm length, F being the force.
(note: supposing that the wing-fuselage angle is 90°, sin teta = sin 90° = 1, thus sin teta factored out)
Now the arm length between the engine and the fuselage is significantly longer than the engines at the root configuration. So for the same torque, the force acting on the fuselage will be alot smaller to compensate for the longer arm. As a result, the left roll moment will be reduced significantly.
Also the torque' s effects will be relative to the size/weight of the airplane.
If you put 2 x 11000hp TP400 turboprop engines on a piper seminole, you're going to have a take-off run like a rocket but if an engine fails, your entire airplane will start rotating around the engine, even on the ground.
If you want to imagine the effects, just compare it to losing a wing during flight, but the same effects happening to you on the ground.
If you put 2 x 180hp O-360 engines on an ATR-72, you could get to move the ATR-72, and if you have an endless runway, you could make it take-up a certain speed. If an engine fails at any point then, it wouldn' t matter because it would have no roll nor yaw effects to the airplane.
The effects can be discussed for hours but in the end they would differ so much depending on the airplanes considered due to several factors such as the actual torque, the distance of the engine on the wing from the CG and other factors like air density.
In general, for twin aircraft of average thrust to weight ratios, due to the lateral moment arm length to the center of gravity on the Y/pitch axis, the effects of such engine/prop torque that ultimately cause a higher drag on the main gears is barely perceivable, due to the arm length reducing the perceivable force on fuselage and landing gear and also because the force would act in a more balanced way on both sides of the main gear compared to a single engine aircraft with an engine's torque axis running between the main gears, segregating the force more unequally onto the left main gear.