The point with the Megane analogy is that the rapid spread of its use is a reflection of the sheer quantity of races and number of players vs. the rate at which PD apply a BOP means that given a choice or option, one outcome will become very dominant very quickly, and this the leads the community to get all antsy about it. And, this is something that is simply not comparable to real-world racing - there are not thousands upon thousands of races taking place between BOP adjustments.
Can't speak for the Gr.3 races, I've only done one Gr.3 daily out of a measly total of sixteen sport mode races (one was a N200? race, the rest have all been in Gr.4)... and in the Gr.3 race, about half the grid was 911's at the start, and the top 5 was 911's. I've no idea if this is still the case, but that race didn't strike me as much different to the Gr.4 situation with the Megane.
I understand the analogy, but the reason I don't think it's helpful is that the game was unbalanced, clearly. If GR3 was in the same situation, you'd see the same number of threads. Instead all I can remember there being (post launch) was one thread asking about the 911 GTR3, in which, most people concluded that it wasn't necessarily the fastest out right.
I think it would be fascinating for them to do from vehicle dynamics simulation standpoint, and it'd be somewhat representative of real-world championships that have tyre-wars... however, I don't believe that given the physics simulation limitations, putting additional variables in would do much for the game - I accept that in the real-world it can make a big difference, but in a limited physics simulation, it simply acts as one variable in a relatively simple equation that ends up having a predictable outcome quite quickly, thanks to the thousands of races taking place each day. From the game point of view, it becomes little more than a nuance, IMHO. With a more detailed tyre simulation model and handling, sure, there'd be more value, more again with a more detailed suspension model, even more again with a more detailed track model, yet more again with a different environmental effects model... ... but personally, and I'm sure others would disagree, whilst accelerated tyre wear has to be a thing, it's less about realism, and more about entertainment anyway (and I'm alright with that btw).
I don't disagree with this, but just because the physics model is relatively basic doesn't mean it would have little/no value.
I also don't think a 'tyre war' type series would be the only application. I think that if the game pushed the idea of tyres being different and having different characteristics, it would help people be better drivers, learn that braking zones aren't binary and that tyre wear is important and can be managed/helped.