From what I understand about X-Plane (without actually having tried it), the flight model is more realistic than in MSFS. Supposedly the flight models are calculated based on airflow over the aircraft, rather than from plugging in a mathematical model that "emulates" an aircraft. This approach would seem to provide more fidelity to the aerodynamic properties of the aircraft model, and would allow a potentially truer simulation of lift, drag, aircraft handling at various speeds, stall characteristics (which MSFS handles pretty woefully), spins etc. Dunno. I'll have to give X-Plane a spin to see if it is any better than MSFS.
It's a different approach really. X-Plane uses something called 'Blade Element Theory'. A shape (a rectangle with an aerofoil cross-section, for example) is divided up into a finite number of sections, the velocity vector for each section is then calculated, using stipulated aerofoil data (if the object has a defined aerofoil) and from this the program can calculate the behaviour of the object.
It works for pretty much anything - wings, propellers, rotor blades etc, and as far as I know is unique in a desktop home simulation. Clearly there are limitations for things like the behaviour of the fuselage and other objects that aren't designed to provide lift (like pylons, struts etc) but short of a full CFD package that is always going to be compromised.
MS uses 'lookup tables', for instance if you were to extract an .air file you would find entries which define strictly how an aerodynamic surface responds to certain parameters, for instance a table that defines Coefficient of Lift Vs Angle of attack.
The advantage of such an approach is that you can 'cook' some very accurate behaviour, such as fuel consumption and lift/drag ratios etc. X-Plane tries to model these things dynamically. It's most definitely worth a look if you're interested.
James