Hi Zen321,
Thank you for very detailed explanation. I have no rejection for your examples. But let me state some points in order to clarify.
I mean by realism, "the real race conditions" , not "the physics rules".

There is no doubt that, additional single gr makes slower the lap time. And according to the tank capacity, the quantity of the fuel taken does not equal to an integer quotient everytime. But we cannot determine this amount by 0.02% in real race conditions precisely. Just think about the fuelman in F1 pitstops. If the fuelman fills the tank longer by 0.14 sec and pulls back the hose 0.27 sec, all the strategy will collapse. The 0.8 sec. gap you stated before (lap based system = percentage system + 0.8 sec.) already includes the human rejection time lag. So, we can say that "the lap based system" seems more realistic.
And the MRT case. It's obvious the weight effect is more prominent in light heavy cars. But in road car and GTR car class (over 1100 kgs + 70 kgs standard LFS pilot) this effect is very slightly. As I know in F1 cars extra 1 liter of fuel causes 0.2 - 0.3 secs lap time loss according to the tracks.
My suggestion is an option such like as km/h - mph or psi - bar options. All I say that, LFS can calculate the fuel quantity instead of us and will be more user friendly. And finally, we say almost same things, but very slight differently.
