The twin fitness landscape properties of neutrality and ruggedness are crucial to the dynamics of evolutionary optimisation. In this paper, we investigate the interplay between these two properties in a complex evolutionary robotics fitness landscape, through the introduction of four robot controller architecture models; the GasNet, uniform, dispersed and plexus models. We show that, in isolation, neither added neutrality or decreased ruggedness (coupling) in the models produces increase in the speed of evolution. However, both effects in conjunction produce a significant increase in the speed of evolution.