I think that the calculation performed by the DashDyno SPD would be simpler than that.
Torque=Force x Distance
In the DashDyno SPD you need to create a profile of the vehicle being tested for power and torque, which requires, gross vehicle weight (including driver/passengers), wheel diameter and gear ratio, and the "force" and "distance" in the above equation can be calculated from those two parameters when combined with the continuously monitored RPM and the time stamp.
In order to standardise the results, the DashDyno SPD also uses vehicle coefficient of drag and frontal area of the vehicle, plus the environmental factors of air temperature, ambient air pressure, humidity and elevation.
Therefore, if you wanted to get real accuracy, to replicate the actual weight of your vehicle then you could modify the vehicle profile before each dyno test to give the correct gross vehicle weight according to the amount of fuel in your tank.
As I see it, an "at the wheels" dyno calculation using a DashDynoSPD will be just as accurate as an "at the wheels" static dyno recording, and in many instances it will be better because there will be little/no wheel slip if the dyno run is performed in the correct gear for the vehicle and all results are standardise and even account for all the losses and aerodynamic effects in real world driving conditions.