The above two posts are correct. But for any given speed my car is running, if I had the choice to run a shorter gear or a taller gear (the portions of the graph where there is an overlap of curves from two or more gears), and my objective was to accelerate as quickly as possible, I would choose to ride on the curve that is sitting highest up on the vertical axis for each given speed.
Take a specific example from the graph, driving at 75 MPH: in 4th gear, I am riding just before the point of peak engine torque. But I would get faster acceleration if I were to instead run in 3rd hear (higher up the vertical thrust axis) even though at that point the engine is actually developing less torque. The gear is doing nothing magical here. The point: at 75 MPH, is the car accelerating faster at the torque peak (4th gear), or is is it possible that the car can actually accelerate quicker at some condition where the engine is actually developing less torque?
It's all a question of frame of reference, and Diesel drivers say torque is almighty because most are conditioned in their (our) driving style to stay in the same gear as much as possible. In that case, sure, more torque equals more acceleration. But if we're in a drag race, we have to put that hubris aside, and row the gears more, which means getting the absolute highest acceleration at any given time; at any give speed at which the car is travelling; and at any gear.
This occurs ONLY AND ONLY at the part of maximizing engine power! End of discussion!