Be interesting to see this thread get into the meat of simulator evaluation and certification ... ?
That's a big subject.
I'd be glad to try and answer questions if you have any.
Whether or not a COTS X-plane package can be used with flight test derived coefficients (or public domain values in the same areas), and provide the necessary accuracy and realism for a regulator to accept the result as adequate substitution for the airplane, I think remains to be seen – although it probably won’t be terribly long before someone does just exactly that and some regulatory authority somewhere around the world will be confronted with making that decision.
You wouldn't gain much from using the X-Plane model, with it's predictive method, because you would then have to try and make it fly to match the flight test data. Easier to do that with traditional aero modelling techniques.
whether a more accurate and more realistic simulation can be achieved through the use of a complex set equations of motion, modified by coefficients derived directly from flight test of the subject airplane – OR – if one can be provided via a set of look-up tables, populated with what someone thinks (or for that matter, has knee-board data to prove) is representative of the airplane under varying sets of circumstances.
You have fallen for the common misconception that MSFS is just look up tables to match observations. There are tables, just as there are in a FFS aero model. These are used to generate some of the coefficients and derivatives. There is a proper, but limited aero model.
Secondly you don't seem to understand aerodynamic modelling. Coefficients are not used to "modify" the equations of motion. Equations of motion come later. First you calculate the aero coefficients and derivatives. From these you calculate forces and moments. Then you insert these values into the EOM.
X-plane allegedly computes coefficients and derivatives from the shape of the 3D model. As the average PC is not yet capable of real time CFD calculations, I suspect this is very simplified and based on published data for aerofoils. Where X-Plane wins over MSFS is that the model they use is more complete, not how they compute the data. Anyway if the aerodynamics of X-Plane is in anyway realistic and non-linear then the only way to do this is with polynomial approximations or function generation (using data tables!).
Several decades ago the simulation industry went to the aerodynamic model; and quite frankly, I think that is why X-plane has a pretty nice product.
Every flight simulation needs an aero model, why does this make X-Plane "nice".