My reasons for filling to the neck were as follows:
-To verify the tank was completely full using different pumps.
-To increase my travel range, and have fewer fill-ups.
-To calculate gas mileage with the least amount of error.
As I said before I realize that there were possible adverse effects to filling up the gas tank completely. However, for the past two years I have had no noticable problems requiring me to change this.
The thermal expansion factor for gasoline is
950X10^-6. Calculate a change in temperature of 5 degrees C for 10L yield a difference in volume of ~142mL. Take that to the extreme from 0-25 degrees C and it's plus 712mL. I doubt this will splash up out of the tank. Not to mention that I only fill-up before going somewhere, not on the way home. Any change in volume with thermal expansion will be used up within the first couple of minutes of driving.
As for the fuel pump vacuum system I hold the pump head just outside of the filler neck and fill the tank by just barely squeezing the pump handle. Of course this is done after I'm sure that I've been grounded along with the pump head and my car. The only thing the fuel pump would be vacuuming up is the vapor which I consider acceptable losses.
With that said, as per the scangauge manual, the fuel economy calculations are affected by the amount of fuel you enter into the scangauge upon fill up. If scangauge is indeed calculating the fuel economy based on injector pulses why would the fuel added to the tank be affecting the outcome?