This study presents a comprehensive comparative analysis of simulation approaches for optimizing the collective motion behavior of robot swarms, with a focus on industrial applications. We investigate the feasibility of bridging the simulation-reality gap by utilizing fast, low-fidelity simulators for parameter evolution, thereby enhancing the efficiency of real-world deployments. Our analysis examines transferability of optimized swarm parameters derived from two low-fidelity simulators to a high-fidelity robot simulation environment. These low-fidelity simulators are employed to fine-tune motion parameters for both ground and aerial vehicle swarms addressing complex coverage problems. The results indicate that while a differential drive simulator requires significantly more time to generate parameters compared to a point-mass simulator, it yields superior swarming behavior and coverage performance across both robot types in high-fidelity simulations. This work highlights the potential for effective parameter transfer to real-world scenarios, paving the way for advanced applications in areas such as autonomous inspection, environmental monitoring, and search-and-rescue operations.