Autonomous vehicles with controllers capable of tracking path and speed in the largest possible operational domain, i.e., beyond and within the stability limits, provide significant safety potential. Dual-motor 4WD electric vehicles (EVs), a configuration already in mass production, can enable speed control when operating beyond the stability limits. This article presents a dynamic inversion-based path/speed tracking control law for this drivetrain topology, solving the nonlinear model inversion in a less computationally troublesome iterative way. The characteristics of vehicle input couplings are exploited to make this unified controller comprehensibly track the desired sideslip only when the sideslip dynamic is unstable. To allow the potentials of 4WD EVs to be fully unleashed, the prioritisation between path-tracking and yaw stabilisation, and the lower bounds of speed when the vehicle operates beyond the limits, are discussed. The framework is then validated in simulation for a reference path with both slowly changing segments and highly dynamic segments.