Time for another round...
All work uses CarTest physics based simulation software which I've posted about and validated vs. documented tests quite extensively. I am going to do some bracketing here with a underrated light version of the car and a non-underrated heavy version.
Let's take the "lighter than 3306 lb" figure as 3305, add 15 lb for either the standard 7 kg of package weight or perhaps just 15 pounds in additional options (not much for options I know this is the light version of the car). As an option which I use CarTest then also adds 68 kg or 150 lb for the driver. In this car I am going to use 450 hp (assuming about 20 hp underrating). This requires 430 ft lb of torque to deliver this a 5500 rpm which appear to be the "knee" of the power curve. I also assume the redline is 7750. For the heavy version of the car I am adding an additional 44 lb for no CSiC brakes and 40 lb for the DCT. I'm also using the substantially less performing 6MT and a lower 7500 rpm redline. I have posted all of the inputs only for the "light/fast" version of the car. I've also shown a E92 M3 M-DCT as the first car and control of sorts in the group.
We arrive at a quite wide range of predictions. Obviously with M-DCT and CSiC brakes I think we will be at the top end of this range. Both missing will substantially hurt maximum achievable performance. At the top end here we are definitely talking Porsche 991S territory or even better as I've stated even prior to today.
In short:
0-60 mph: 3.4-3.8 s (sub 4.0 s is pretty well unquestionable)
0-100 mph: 8.0-8.9 s
60 ft: 1.7-1.8 s
1/4 mi: 11.6-12.1 s @ 115-119 mph (std. US drag racing with 1' rollout)
60-130 mph: 10.4-12.2 s
And for our EU friends
0-100 kph: 3.5-3.9 s
0-200 kph: 12.4-14.0 s
0-1000 m: 21.2-22.1 s
Not that it comes out of CarTest but it is informed by this work, my Nurburgring Nordeschliefe lap time prediction for the car is 7:4X. I just don't think it can match the 991S on the Ring based on weight distribution, size and aerodynamics.
I know everyone wants "THE" or "A" performance number. There is simply too much still unknown to do any better than this type of bracketing analysis. Either way its going to be pretty darn fast, strip and track.