What is realistic?
I think they all are. Maybe.
Do you want to time your car while putting the car in gear, holding the brake with your left foot and preload the transmission to about 1000 RPm with your right foot and floor it and lift the brake synultaneously? I think that will get you "about" 7.5 secs depending on gas used, altitude, Barometric pressure, temperature and humidity, track temperature, track surface, tire temperature, Tire pressure, and probably some more things that escape me at the moment.
Do you want to do the same as above but preload the tranny at maybe 5000 rpm (Some experimantation is required)? This will insure that the torque converter is llocked and if you don't burn things up, you will proabably set the fastest time.
This is likely the only way you are going to get 7.0 secs.
Plus it has to be an SL FWD American (not Canadian) model with a very small man or even a slim girl driving.
It's the weight you know.
The major difference is often in testing techniques.
And sometimes the cars coming off the assy line actually vary.
A Lot.
Sometimes it is based on When they are manufactured/tested.
Twice now, people have tested Mazda pre production cars (Miata and now the RX8) and changes were made that reduced the HP by the time the car got in production. If you bought an early RX8, Maxda will give you back your money.
The HP difference between pre-production and production was significant.
And there are changes rolled into production. Not all changes are done at model year.
Who's to know that the current MO is faster than the early MO?
Or the other way around?
Maybe a change in the CVT profile?
I like to think of the MO as an 8.0 second machine.
IMO the hotrodding of the MO will concentrate on the ECU and the CVT.
Homer