Trelinski,
i drive a 5spd 190e2.6. I rarely go east of Ajax, and park at the Ajax Go Train station since i work downtown. (But sometimes i go to Oshawa to exercise my car and explore the M103's torque characteristics in 3rd gear on the onramps )

The concept of torque is quite abstract, but i have the following electrical analogy: torque is like voltage, and RPM is like current. Multiplying the voltage by the current, you have power (watt = volt-ampere). Except in an engine, both the torque and the RPM vary.

A flat torque curve is the ideal one - having constant torque across the entire RPM spectrum. In this case, the power curve will be linear with RPM.

But torque in an internal combustion engine is influenced by a number of factors across the RPM range, and therefore it is exactly that - a curve. The realistic objective is therefore to have as much torque as possible as low as possible in the RPM range above idle, and to maintain this across the "sweet spot" of the designed engine operating environment. For typical daily driving, this is around 2000-4000rpm. Having at least 80% of the max torque in this sweet spot, for example, makes the power "usable" for daily driving - you don't have to rev the engine high to get torque. Designers resort to variable valve timing and variable induction, among others, to "flatten" the torque curve.

A racing engine on the other hand, is designed to make full use of its redline (they say that RPMs are free). The power and torque peaks are typically high in the RPM range, close to each other, with little attention paid to low-rpm torque (look at the Honda S2000 figures). The induction system, exhaust system and valvetrain (e.g. VTEC) are all tuned to maintain the torque high up in the rpm range. Close-ratio transmissions are typically used, so that the RPM stays within the sweet spot when you upshift or downshift.

i hope this did not make things more confusing.

may the torque be with you ...