Autonomous navigation of drones in environments with intermittently poor GNSS reception remains a complex problem with small and lightweight sensors. Conventional kinematic frameworks, fusing inertial sensors with satellite positioning, often falter under such conditions. Besides, recent developments propose the incorporation of flight aerodynamics through a so-called Vehicle Dynamic Model (VDM) to mitigate position drift during GNSS outages; however, these approaches have predominantly employed a single Inertial Measurement Unit (IMU). In this research, we improve navigation resilience by proposing a novel architecture that fuses measurements from multiple IMUs with GNSS and aerodynamics via the VDM-based system. We rigorously validate this architecture through four comprehensive flight tests, each comprising two GNSS outages lasting two minutes. The results indicate statistically comparable performance when fusing data from one to four small IMUs with only a modest increase in computational burden in the latter case. On the other hand, the multi-IMU system exhibits considerably higher information redundancy and therefore improves the robustness in terms of maintaining effective drone positioning even during multiple IMU failures and GNSS denial, assuming fault detection and isolation are correctly addressed which are outside the scope of this contribution.