This study presents the vehicle control optimization of a Formula SAE (FSAE) electric vehicle developed by National Taiwan University Racing Team (NTU Racing), utilizing a dual-axle dynamometer and a real-time Hardware-in-the-Loop platform from Chroma. The novelty of this work lies in the comprehensive system-level validation of independent torque control strategies, namely Torque Vectoring (TV) and Traction Control (TC), implemented directly within the vehicle control unit (VCU), and the high-fidelity simulation of dynamic driving scenarios based on the FSAE circuit. The vehicle features an independently controlled rear-axle, two-wheel drive (2WD) configuration, consisting of two in-wheel motors, self-developed inverters, and planetary gearboxes. During testing, a pre-built CarSim driver model provides throttle, brake, and steering inputs to the VCU via Controller Area Network (CAN) interface. The VCU, in turn, computes the independent torque commands according to the TV and TC strategies, which are then transmitted to the inverters and applied to the motors. The resulting torque output from the planetary gearboxes is measured and fed back into the CarSim vehicle model to simulate the rear wheel dynamics and command the dynamometers at the corresponding rotational speeds. The results show that with the dual-axle platform, the independent torque control strategies could be tuned effectively to improve vehicle dynamics, offering a more quantitative and precise approach for performance optimization compared to conventional Model-in-the-Loop (MiL) evaluations or driver-dependent feedback from track testing.