The modelling representation of windshears in simulators has the same basis as that of the aircraft - measured data. But while the aircraft model is derived from aircraft data, the windshears (and indeed other things) are dervived from more general data. There are in fact a range of "canned" wind profiles derived from a number of real life events, and those can be used (they are also used in windshear warning/guidance certification tests).
I would be very cautious in assuming that the assumptions on which the aircraft performance calculations are based are themselves regulatory, especially in the operational world - just because Part 25 says to assume something doesn't mean part 121 (say) enforces it. I would also be very cautious about extrapolating those same performance assumptions outside their intended scope; there's nothing in part 25 about assuming any kind of windshear for the purpose of performance calculations - the closest there is would be the 0.5x/1.5x factor on STEADY wind, but that's hardly applicable to the shear case.