Nice try. Dynojets directly measure power at the wheels. They then back calculate against the RPM of the motor. On something like a dynapack that measures torque at the wheels directly, this is true. That's why they make you punch in the overall gear ratio so the computer can divide it out, then back calculate HP.

Horsepower at the wheels is independant of gear ratio or tire diameter. The pitfall of your logic is the 3rd term in the tq-hp equation. That's rpm. If you jack up the wheel torque with a higher final, you're also spinning that wheel faster per engine rpm. The 2 factors are equal and cancel out.