gary: "The way I used a DVOM to measure fuel was to measure the resistance of the sending unit"
OK, very interesting, thanks for explaining. I never would have thought of that. You had to be very patient to collect all those calibration readings. And then when you take subsequent readings, you have to make sure the car is on a level surface, and you have to give the fuel a few moments to settle in the tank (stop sloshing around). I'm sure you know this.
Also, couldn't the float mechanism have some significant hysteresis? That is, it might not respond in a smooth, accurate way to small changes in level. If you drove a relatively short distance, and are therefore measuring a very small change in fuel level, then the error could be significant, I suppose.
Then again, when you did your calibration procedure, you may have observed the float responding in a smooth and sensitive manner. So I could be all wet!
Anyway, I wish there was simple, quick, accurate way to measure fuel level, on an ongoing basis. Waiting for the next fill takes too long, and it's also not very accurate (until you average a bunch of tanks together).
You probably realize the pros use a small bottle, and they weigh it, before and after. Too bad we can't do something like that.
__________________
|