Ignoring the issue of the absolute amount of heat, here's some other background that might help clear up a few things...
For a resistor, we're always talking about the wattage converted into heat. That heat goes somewhere. It starts in the conductor (wire), but as you continue to apply electrical current and generate more heat, the heat conducts or radiates out of the wire. If the heat is generated faster than the conductor can dissipate it, it builds up in the conductor causing its temperature to increase. If this condition continues, at some point the temperature will become high enough to melt or sublimate the conductor (or part of it) and open the circuit.
Leaving aside certain issues such as choice of materials and wetting/wicking properties... The rate of heat transfer is dependent on the difference in temperature, and the area through which the heat flows, and it's inversely related to the length of material through which the heat has to flow (in this case, the thickness of the wire). A larger temperature difference causes heat to flow faster. A larger surface or contact area causes heat to flow faster. A thicker wire causes heat to flow more slowly.
If the goal is to transfer as much heat as possible as quickly as possible (i.e. to produce the most vapor in a given period of time), the most efficient conductor will heat quickly, maximize the surface area, and minimize the wire thickness. This means using the longest, thinnest wire possible, while still accounting for the actual values for the materials (i.e. liquid and nichrome). All of these optimization (hot, thin, & long) also increase electrical resistance (or for heat, depend on increased resistance). Higher resistance requires a larger voltage to convert the same amount of energy to heat.
So it seems to me that if you want to design an atty to produce maximum vapor in an efficient manner, you end up with a high resistance device, and that requires the use of higher voltages to generate a given quantity of heat.
There's a lot of other factors, such as whether heat flows faster through the conductor or across the conductor-liquid boundary, how long you apply the electrical current, the wicking system that replaces evaporated coolant; and other heat losses such as through the physical wire mount, heating of the liquid without achieving evaporation, and radiation through the liquid or directly to air. I dare to guess that we don't want to heat the liquid unless we're actually evaporating it, which argues for very fast heat transfer, which happens at very high temperatures. As anyone that solders knows, too low a temperature applied for too long a time is what burns things unintentionally. Conversely, evaporating the liquid too quickly will cause a space to appear between the liquid and conductor, potentially burning the atty. Undesirable chemical processes might also occur at high temperatures--I really don't know what exactly happens when you "cook" the liquid either at low or high temperatures. Obviously, these things have to be factored into the design of an atty, or we'd all just use very high voltages and the longest conductor possible. In the end, it's all about making tradeoffs.
