Automatic numerical optimization is playing an increasingly central role in aircraft development. In previous work, the authors demonstrated the feasibility of high-fidelity dataset-based shape optimization for aircraft design. While aerodynamics is captured using high-fidelity Reynolds-Averaged Navier-Stokes (RANS) computational fluid dynamics (CFD) simulations, other disciplines such as mass, structure and flight mechanics are included at a lower fidelity level, reflecting a pragmatic multi-disciplinary approach. The current work introduces a high-fidelity treatment and sub-level optimization of the key aircraft structure, the wing, into the shape optimization loop. The high-fidelity aerodynamic dataset serves detailed sets of rigid load conditions covering the entire flight envelope, which are mapped onto a high-fidelity finite-element (FE) Model, allowing for efficient and accurate structural analysis and sizing. No a-priori assumptions must be made regarding which regions of the envelope are decisive for the aircraft structure sizing. This reduces structural design risks significantly and enables robust shape optimizations with co-optimized structural members, reducing the uncertainty in the mass estimation, and thereby increasing the accuracy during the mission level evaluation of flight performance as well as stability and control. Effects and implications introduced by the unidirectional coupling of the aerodynamic envelope loads onto the structural members during the optimization are analyzed and discussed.