I agree with Yasin's point. To add to his views, the cold flow analysis can be used as a tool to make all the geometrical changes to the component being designed without any reactions within the system. Cold flow analysis can be validated with experiments.
Classicaly, experimental cold flow studies (mean and rms PIV fields for example) are taken as a first validation for CFD simulations of a complex system. It helps to develop accurate combustion simulations by assuring that the aerodynamics are already well modeled
I fully agree with previous answers. I can add one more thing. The combustion process of a hybrid rocket motor is diffusive-limited. In this case, the fluid dynamic structure of the flame is not alter by the presence of the flame itself (check Poinsot-Veynante textbook). The flow path that you find in the cold flow tests are very likely to be the same that you find during combustion, if the blowing from the burning surface is small enough (and in general it is, for pyrolyzing fuels).