This work examines the impact of diffusion dynamics on community-aware centrality performance on networks with varying community structures. We extensively analyze the effectiveness of multiple centrality measures in ranking influential nodes by comparing four different network diffusion models (Susceptible-Infected, Susceptible-Infected-Recovered, Linear Threshold, and Independent Cascade models) on synthetic and real-world networks. Results show that the community structure strength, the internal dynamics of the diffusion model, and budget availability are critical to determining the diffusive power of nodes. Simple contagion models (SI, SIR, IC) show very consistent patterns of diffusion, while the complex contagion dynamics provided by the LT model behave quite differently. This study sheds light on selecting the most influential nodes to maximize diffusion in real-world applications in epidemiology, analysis of social networks, and marketing.