[RC-Monster Mike]

The tooth width variance pointed out by Brian is called the pressure angle - Traxxas Revo spurs use a 14.5 degree pressure angle, as do my standard mod1 pinions with 12-15 teeth. The 16,18 and 20 tooth mod1 pinions that I make use the stronger 20 degree pressure angle(same as the clutch bells on most nitro vehicles). The 14.5 degree pressure angle has slightly less rolling resistance, while the 20 degree pressure angle(which is the "standard" for most gears) has slightly stronger teeth. My hardened pinions were made from real live clutch bells - I just chose a lighter weight method vs. FD.
The problem is, there is a wide spectrum of quality on the clutch bells, and mesh is probably the largest factor for the gear's survival. As Dustin pointed out, the electric conversions also "use" the gears much more harshly. Nitro vehicles also have a clutch, which effectively limits the torque transfer between the gears, along with not using the gears for braking. Also, a typical electric motor is suspended using only 2 screws at one end, with the motor hanging off the other end - most motor mounts used for this also exhibit quite a bit of flex(landing a jump, for example), which undermines the gear mesh and accelerates wear substantially. A nitro motor, conversely, typically is mounted to a flat chassis with 4 screws and on a flat plate/mount, which holds its mesh MUCH more rigidly. The high stress of a brushless sytem, along with the lack of any clutch/slip and a relatively flexible motor mount typically leads to the demise of these gears. Metal on metal is far from ideal in the real world anyways. To illustrate the importance of a rigid gear mesh, and the effect it has on wear, consider the hardened idler gears used in the emaxx tranny - these gears last a very long time, even though they mesh against a hardened top gear. The largest factor is the RIGID mesh. Both gears are supported on both ends and the mesh doesn't change in use, which results in much longer lasting gears(metal on metal will always wear, just much more slowly when proper and rigid gear mesh is present).
Nylon and delrin(common materials for plastic spur gears) are plenty strong for the application, along with having much higher natural lubricity(less rolling resistance), and less rotating mass. They also have the benefit of allowing almost any material to be used for the pinion, resulting in a less expensive and wider availability of gearing. Maintaining the mesh is still important, of course, but a metal pinion on a plastic spur is much closer to ideal in many respects.
I still have quite a few hardened pinions, but I took them out of inventory simply because I was not thrilled with the consistency(varied quality mentioned earlier)/quality of the product. In short, I would rather lose money than my reputation and I wasn't satisfied with the product in the end. they were made from clutch bells, but in harsh situations, they still failed occasionally - not up to my standards. This is a major reason I developed the slipperential setup, and will adapt it to different applications. The performance of a center diff, with the durability/reliabilty of a slipper. My personal opinion is that direct drive without a slipper should be avoided. Sure, it will work for a while, but without the outlet for extreme stress relief(slipper), the diffs take a beating, as does the motor and speed controller(high amp spikes at the controller, and eventual dislocation of the motor shaft/rotor).