So Jimmy Broatbent just tested AMS2
His conclusion at the end is direct and simple: it's not yet the simulation he was waiting for and said at 13:00 'AC looks better'
He pointed out a lot of weaknesses in the car behaviour and that the car doesn't really feel like a DPI car to be honest.
Don't forget Jimmy has a high expertise in racing simulation but he's now a real race driver.
I liked Jimmy's video and the behaviour he describes could indeed be something that needs ironing out (I am having a lot of fun with GTEs haven't tried the DPi yet) but bear in mind a couple of things. Firstly, it could be something wrong with the driving model and in that case, Reiza needs to fix it.
Another reason is a setup issue. Car might be too soft on the rear for instance.
He has a different throttle application than me in corners so with the default setup, I don't know if I would suffer from the same. He recognizes that this behavior happens when he pushes the throttle a bit early while cornering and that he suffers a four wheel slide while he thinks the car should be understeering (around 8 minute mark). Pushing the throttle early will shift the weight of the car back quite a bit. He has real life experience with a car kind of similar to this but with about half the power so the weight shift that happens will be much smaller. This could be a simple setup away from being fixed. It could be an indicator that the car has the center of balance (?) too far back. If you have too little rake, too soft a rear or too stiff a front (or any combination) pushing the throttle will change the aero profile and weight distribution of the car causing the mass of the car to shift back further with throttle input. Too stiff a diff (and I tend to find AMS2 default diffs way too stiff) will also add to this behaviour and can further compound this issue. Also, his front tires are not really up to temperature across his race which is another indicator that the car balance is slightly too far to the rear (in the Vette, I open up the diff, turn BB forward a couple of clicks and soften the front ARB quite a bit in order to get equivalent pressures and temperatures everywhere).
Jimmy also comments that when he was a bit softer with throttle application (8:40) the car responds better and those weird slides don't happen which to me points towards his driving style suiting a more relaxed diff and probably a more front balanced car so that the weight keeps more at the front with increased throttle application translating in not having as much of a loss of grip in the rear.
Again, it could be a flaw in the physics mind you but it isn't necessarily that.
For instance, he compares the car to iRacing LMP2s. iRacing has a much more simple aero model for instance and that could be why they feel more planted (the more complicated the model, the more it has where to fail). So, take Jimmy's opinion for a slightly better than average opinion but he doesn't really attempt to get rid of the effects he doesn't like via setup changes. I understand his time is limited and he has other things to do which is fair. And it is another data point I am sure Reiza is keeping an eye on.
EDIT: Emptybox posted a two hour video on the DPI and he mentions how very much the same it feels to the iracing one. Emptybox has a smoother driving style compared to Jimmy so there is that.
Edited the post in order for it to make sense. I just got my covid jab and my mind is a bit foggy atm.