Yes of course making any changes to a stock car technically makes it a theoretical car but all of those changes are well known to the world. It's well known what happens to a car when you fit a racing exhaust or a racing gearbox or add 500bhp and so on. Yes it's still a rough estimation, but it's based on reality.
This is no different from case with the X1 or other theoretical cars that are designed to be realistic.
It's not proven at all what the X1 as a whole would perform like, no car exactly like it has ever existed in reality.
That's not really an issue. It's a physics engine's job to figure out what the car is going to do, and whether a real X1 or not exists, the physics engine should provide the same answer. A real Viper ACR exists, it's nothing like the one in GT5, but that does not make the concept of the ACR unrealistic. It just tells you that GT5's physics engine (and in this particular case, the parameters used to model the vehicle) are off.
GT5 is a simulator. It does not care what exists in reality and what doesn't. This is because it's a simulator. We can add arbitrary rules to the simulator, like it should only simulate things that have been built in real life and tested to 99% of their performance envelopes, but that has nothing to do with simulators by default.
Now yes, if you created a world class simulation as F1 designers use and calculated all of it's potential performance you might be able to show exactly how the theoretical X1 would perform. But remember even the best simulators aren't accurate to the real world, you couldn't be absolutely sure until you built it. Just the same as Adrian Newey isn't certain how his new car/parts will behave when the car takes to the track.
And GT5 would give you the same car. Just like it fails to simulate so many real cars properly because its engine can only do so much. In essence, whatever error there is in Red Bull's or PD's or whoever's predictions on the X1's performance could very well be within a narrower tolerance than the error introduced by GT5's sometimes primitive physics. Building a real X1 won't get you a better X1 in GT5 because the X1 in GT5 is a result of the GT5 physics engine. You need a better physics engine to get a better virtual X1. This is true no matter how much is known or unknown about X1's in reality.
Will the new physics engine be inaccurate? Yes to some degree. Would we need to build a real X1 to be sure that the virtual car is realistic? No. Models are sufficient to imitate a real car. If they weren't, no one would use simulations at all.
An argument against the X1 needs to come from outside the game. Either enforce "simulate what is real only" or question the methods used by RB and PD to come up with the design of the car.
GT as a simulator in the sense of plugging data figures into it's physics engine is very, very flawed (Mainly with tyres and aero) so no, I argue that the performance of the X1 in GT is not like reality at all.
Then why are you arguing against the X1? If that's your stance you should argue against every car in the game, because what you said really has nothing to do with the X1 in particular. You're right, it's the physics's engine's fault. That means the X1 is as accurate as every car that the physics engine simulates.