There's another problem with the belite fuel sensor. It works on the pressure (or weight) of fuel, which varies depending on fuel type. If you always use mogas or avgas, then no worries, but if you swap between the two depending on availability, then you'll never be sure. I guess in that case you could calibrate with the heavier fuel then the lighter fuel will under-read - ie there will be more fuel than indicated which has to be better than the alternative...
Hmmm... Here's a thought. You know if you blow across the top of a bottle, then the sound changes depending on how much fluid is in it. Likewise if you tap a glass. The resonant frequency of the container changes with the volume of the contents. So, if you set up a sensor to measure the resonant frequency of a rigid tank (wouldn't work on bladders), that should give you a measurement that is independent of both the fuel type and tank shape, and it should be independent of contaminants too. Just like the Belite pressure sensor, the instantaneous reading would vary wildly, but filtered with a time constant of something like a minute, it should read very well. I wonder??