Remember that ultimately to address one of main tasks in the scenario letter, you need to address whether the responses of the proportional controllers you designed were first order or not. Probably the easiest way to assess this is graphically when you are calculating the closed-loop time constant using the log plot discussed in the lecture slides (I discussed this with a number of people today in my office hours). That plot will only make a nice straight line for the calculation of tau if it is a true 1st order response...otherwise, it will be tough to calculate a tau, because after all, the time constant is only appropriate for a first order response, and a system that shows a curve or changing slope on that figure is not first order. You don't need to show this analysis for every proportional gain, just perhaps give on example of a first order response and one that is not, or at least mention this is the method you used to assess model validity if you are short on space.