Physical parameters of the constituent modules and gait parameters affect the overall performance of snake-inspired robots. Hence, a system-level optimization model needs to concurrently optimize the module parameters and the gait. Incorporating a physics-based model of rectilinear gaits in the system-level optimization model is a computationally challenging problem. This paper presents a case study to illustrate how metamodels of the precomputed optimal rectilinear gaits can be utilized to reduce the complexity of the system-level optimization model. An example is presented to illustrate the importance of concurrently optimizing the module parameters and the gait to obtain the optimal performance for a given mission.