I thought the whole point of the exercise was that the conveyor matched the speed of the plane exactly, so that it was travelling at 0 mph relative to the ground, even if the engines were moving fast enough to pull the plane through the air at 75 mph. Given that the plane doesn’t move forward, its wings don’t grab air, it can’t rise. My scenario just makes plain that it isn’t moving relative to the ground, hence no rise.
The point is, that barring extraordinary circumstances, no matter what the conveyor belt does, it can’t stop the plane from moving. It can make the wheels spin very fast, but (except for those extraordinary cases), the plane doesn’t care how fast its wheels are spinning. So the plane just goes ahead and moves forward relative to the air and to the ground and to everything but the treadmill, just as if the treadmill were stopped.
What Chronos said.
It should be added, however, that this violates the statement of the problem (if you’re assuming the problem calls for the treadmill to hold the plane motionless with respect to the ground). So if you disallow “extraordinary circumstances” (and nothing wrong with that), the result is that it’s impossible to satisfy the original problem.
Assuming we’re talking about the reformulated statement from the first post, I tried to write the statement in such a way that it doesn’t say “the conveyor keeps the plane stationary relative to the ground,” but rather, “The conveyor…is designed to [keep the plane stationary]…”
I felt like saying that it was “designed to” keep the airplane stationary was OK, because the trick is that it doesn’t matter what it was “designed to” do, it can’t do it. There was a massive design failure. 
We could tweak it this way:
“The conveyor is computer controlled, and will always try to keep the airplane stationary relative to the ground around the conveyor belt. So if the plane starts to move forward down the conveyor/runway, the conveyor will immediately spin in the opposite direction until the plane has returned to its original location relative to the ground.”
The weasel word “try” in there make me worry that people who hear the problem will immediately say, “Well, can it keep the airplane stationary or not?” Which is a problem, since in order to be deceptive the problem has to imply that the conveyor is capable of keeping the airplane stationary relative to the ground, without actually saying so directly.
I gotta say, though, I’m sympathetic to butzi64’s point. Most people who hear the original problem are not thrown by its little ambiguities and whatnot - they’re led astray by the big gaping hole in their intuition, which is the point of the puzzle. So maybe this is all just nit-picking.
Don