As for the gap, it's always the same thing. I have already warned against manufacturers using simulated lap times as bases for their claims countless times. You can put just any data you want into your simulator and pretend those are the same as in the actual car, but really, you have no way of knowing that until you build it and until you run it. In other words, you have no way of knowing what the correlation between the simulator and the real world data is until you have the real world data. The simulator might have a good correlation to some existing car - but that's a car where you already have the real world data and where you had already fed it to the simulator to get that correlation. I guess that under the goading of management/marketing people that try to make you state the most optimistic, outrageous prediction possible, or just lost in their own hype, some engineers tend to forget that and operate on hopium that this time even their most optimistic assumptions will turn out to be true. And it's the same thing with the driver. It's well known (and really quite obvious) that in a simulator you will be able to achieve a much better time than in reality, just because you can drive without the fear of crashing and take every corner going 100%. But then it's only too easy to imagine that your driver will somehow will be able to recreate that in reality if he feels really brave, lucky, and everything just comes together. Of course, it never does, but even then, you can just pretend that that has to do with the driver and not the car, which is definitely capable of that fantastical time you got in a simulator.
