This investigation analyses the 3D performance of two optimal front wing designs according to the Federation Internationale de l’Automobile 2021 regulations. These regulations imply a 3D five-element front wing, with ground effect and more than 30 design parameters and geometry constrains in practice, which has not been studied before in the literature. The two optimal front wing designs were obtained by parametric optimisation with sampled Kriging interpolation surrogates and a posteriori adjoint-based design refinement. The optimal front wings have been tested on 3D Computational Fluid Dynamic simulations to analyse the generated downforce and flow interaction with the front wheels and brake ducts, as well as the reproduction of influential vortices. Although the parametric design achieved a slightly better downforce-to-drag ratio, the adjoint-based design exhibited a greater ground suction effect, lower pressure distribution on the tire treads, and some deviation of the Y250 vortices to the wheel, which affected the flow over the wheel&brake ducts. This work demonstrated that quick improvements can be achieved starting from 2D analysis and running 3D simulations of relevant parts of the car, which is unaffordable to date for direct 3D simulation in a design space reduced to 17 dimensions and constrained by regulations.