A simple metaphor about this issue: LFS tyre modelling looks like such a electric kettle.
A quick heating in 5 minutes and finally boiling point. However we need "steam locomotive" system. After the starting temperature level the steamboiler of locomotive must be fed carefully. Over coaling will blow up the steamboiler and inadequate coaling will reduce the temperature.
Joking apart, too many coefficients are in the calculation of tyre modelling.
As I know;
-Tyre model (slick, normal, etc.)
-Tyre compound (soft, hard, etc.)
-Tyre pressure
-Track surface (smoothness and vise versa)
-Track design and ideal race line
-Weather conditions
-Track surface temperature due to weather conditions
-Car type and class (FWD, RWD, torque, etc.)
-Total weight of car (including pilot, fuel)
-Weight distribution of car
-Car setups (almost all of them)
-Pilotage
-G force
-....
So, all coefficients above affect tyre modelling synchronously. Of course their influences differ from race to race. Some of them are dominant in a race specifically, although affect in another race slightly.
Another dilemma.. Does tyre wearing problem slow down the laptimes? Not all the time. (i.e in F1, the best laptimes are reached when fueltanks are nearly empty despite the tyres weared out. Herein weight effect is bigger than tyre wearing efffect.)
Finally, in order to alter this complexity the tyre company advisors for race teams should advise the devs, if possible.