How can FOV alone play tricks as all it does it changes the viewing angle without altering the position of the view point and thus perspective?
For example let’s compare the 1st(DCS) and 2nd(RL) shots in my original post (RAZBAM AV-8B Harrier II): you can tell that the point of view is closer to the HUD in the first shot despite it apeering to be otherwise. How can you figure that? Knowing that the HUD image is collimated (one of it’s features is that it*s apparent size is always the same) you can actually judge that the FOV in the 2nd shot is narrower than in the 1st shot. Notice that the HUD symbology fits the HUD frame in 1st shot while it doesn’t in the 2nd shot. That implies that the 2nd shot was taken further away from the HUD glass than the 1st shot. So maybe the the cockpit model is correct but untill we can test it ourselves and compare it to RL pictures we wont know for sure.
Just to ilustrate my FOV point here are few shots from CAP2.
First two taken from the same position with different FOV (see over glare shield)
Now with view point closer to the cockpit (see under glare shield)
Between changing FOV you can’t change perspective to see under the glare shield.
I can’t say that the CAP2 cockpit is modelled properly either, you would have to laser scan the thing to be perfectly sure I guess. Even if you had technical drawings you can’t be sure you’ve put all the planes at correct angles in 3d space or was the documentation thorough enough so you may make that claim?