I feel the need to add some more. Those who are saying that the plane would not take off are basing that on reading the problem as defining the plane to be stationary. Since the question is poorly worded and barring a clarification from the writer, it is possible to interpret the question that way. However, the problem with doing that is two-fold: First, if the problem simply reduces to: "If a plane is stationary, will it take off?", why even bother asking the question? It's so basic that it's pointless. Second, just defining that the plane is stationary does not address the question of whether such a thing is physically possible. And while we engineers are arguing about what is physically possible, those who are saying the plane is stationary are simply assuming it without basis in physical reality. I think if you analyze the problem - think through the steps of how it would work - you'd find that it would be useless to ask the question you are describing for the two reasons above.
However, if you guys do think that your position is physically possible, please follow the steps of how it would happen and explain it. If you do that, I think you will find the scenario falls apart. Let me start it:
Step 1: A plane is sitting stationary on a stationary treadmill.
Step 2: The plane fires up it's engine and begins to accelerate.
Step 3: The treadmill senses the motion/acceleration.
-------Step 3 is a toughie: how can the treadmill sense the acceleration? Unlike a car, which exerts a direct force on the treadmill in order to accelerate, the plane does not. About all you could do is sense the motion with sensors along the surface of the treadmill.
Step 4: The treadmill responds to the motion of the plane and begins to move...how fast?
-----Step 4 is where the scenario completely falls apart. Since the treadmill is not capable of exerting a direct force on the plane in the way that it can on a car, it is not possible for the treadmill to instantly react to it's motion and keep it stationary like it can a car. It could keep a car stationary regardless of what the car does - accelerate, decelerate, whatever. With the plane, the only way the treadmill can react is by speeding up - and while the treadmill accelerates, the plane is still accelerating with respect to the ground. The treadmill can't stop the plane and keep it stationary in a stable situation like it can with the car.
The only physically possible way to stop the plane is to accelerate the wheels until the wheels actually fail - heat up, burn up, disintegrate, ripping the landing gear off the plane. If the problem was meant to allow that, fine - you could always make higher speed wheels and then the problem becomes a battle between the engineering of the wheels and the acceleration capability of the treadmill. But again, that's not a very useful discussion. Plus, the scenario is uncontrolled - the treadmill is not under any sort of active control - it's not reacting to anything the plane is doing, you just turn on the treadmill when the plane fires up it's engine and see who wins the battle.