Feedback

You've taken the idea of a theme on board from the beginning, and that's promising. What I'm not quite so sure about is whether the choice of modelling exercise is as well-aligned as it might be to the evaluation you aim to make (if any specific model-building exercise can be!). You need to bear in mind the generality of things that get to be built in software engineering when you consider 'why EM is distinct from other forms of software development': maybe your case study is really oriented towards a rather narrower comparison (e.g.) between object-oriented and empirical modelling in simulation. Good to have identified Paul Ness's thesis as a source, as I think this has some peculiarly relevant ideas. There is more recent stuff on the EM-OO comparison (e.g) in the EM theses of Sun, Heron, Wong, Wang that is probably less relevant if you stick to the narrower comparison (but see e.g. Heron's cruise control and Wong's dishwasher - in ~empublic/projects/javavccsHeron2002 and dishwasherWong2002).

You may find some material in ~wmb/public/projects/games/CRICKET that bears on the Frisbee simulation. It would also be rather good to find out more about the thinking behind Simula, the first OO language, specifically designed for simulation I believe, though detailed documentation of this has proved rather elusive. (Good to bear in mind that OO as it applies today in software development practice is rather - if not radically - different in its orientation!)

All in all, then, I would advise that you focus on developing your model and contrasting it with what you could do within the object-oriented idiom (where some of the general thinking explored by Heron and Wong could be helpful sources of inspiration). A good model is worth a lot by itself, so I don't think there's any problem about shifting the balance towards the practical side if necessary. Harder to move towards emphasising the paper without good experience from the model though.