As a result of all this recording and analysis, do the figures stack up against the trip computer? Or do they confirm that the trip computer consistently paints a rosier-than-real picture about average fuel consumption, while at the same time being a panic merchant about the need to fill up?
I usually run the bowser figure through a calculator when I fill up and check it against the average shown on the trip computer (D4 3.0 with all known updates completed). The trip computer is consistently about 10% lower than real consumption: when the trip computer gives an average of 9.1, the bowser shows actual consumption of 10l/100km; computer 10 = bowser 11.1; computer 11.1 = bowser 12.3, etc. I seldom fill up at the same place, so it's unlikely to be a fault with the bowser.
The average fuel consumption figure shown by the trip computer also never tallies with the distance to empty, which normally warns of certain death by about 700km even when the average consumption figure is showing around 9 so there should be a good 900km in the tank (or 840km once you factor in the error in the average). I tend to err on the side of safety and fill up by about 700km but I'd hate to need an accurate figure to wring maximum distance out of the tank.
This is only the 2nd car I've had with a trip computer and the other one played similar games (Peugeot 406 diesel), so perhaps all trip computers do this? Or is it just mine?

