Rapid sensory feedback is necessary for executing complex movements with precision. In many flying insects, strain-sensitive structures on the wings provide an essential component of this feedback. Structural properties (e.g. wing geometry, flexural stiffness) interact with forces acting on the wing to produce the local strain sensed by the nervous system. While previous work has examined how wing structure determines aerodynamic performance, the impact of wing mechanics on strain sensing remains unexplored. Using a computational model inspired by the wings of the hawkmoth Manduca sexta, we examine how strain-sensitive sensors can be most efficiently placed over a flapping wing to detect body rotations via Coriolis force-induced strains on the wing. We modeled the transformation from strain to action potentials (spikes) using a linear filter and a nonlinear function, which were derived from prior experimental analyses of strain sensing in Manduca. We show that spiking sensors, conveying a signal that is both spatially and temporally sparse, can accurately detect body rotation over a wide range of wing stiffness values: typically only five sensors are needed to achieve near-peak accuracy, in many cases greater than 90%. Structural and neural encoding properties interact to jointly determine the optimal number of sensors and their locations, concentrated at either the wing base or the wing tip. Moreover, sensing performance is robust to both external disturbances and sensor loss. Our results show that small amplitude, dynamic inputs can be extracted with spatially and temporally sparse sensors in the context of flight and point to the importance of the joint evolution of structural and neural encoding properties.