Understand that LiPo battery SOC (state of charge) determination is a bit tricky; there’s a balancing act between maximum discharge time / allowable depth of discharge and battery health, in other words, where is “zero”, since you never actually discharge any Lithium chemistry battery “all the way”, or it’s Done Forever. So device manufacturers have to pick a minimum which is a “reasonable” point to disallow further discharge. So there is an offset to, and also some non-linearity in, the discharge curve, and that has to be matched up to a percentage number to display to the user.
A good analogy is vehicle fuel tank transducers, most people can relate - who hasn’t had a car that would go, say 250 miles from Full to 1/2 but then go 300+ miles from 1/2 to Empty (or worse, vice versa)?
Getting it tweaked takes some iteration and a good sample size of batteries, to accomodate tolerances, variation etc. Anyone who has done best-fit curve-fitting in software will appreciate this (ahhh polynomials…)
Overall, I think Sansa has done a Fine Job with this in the current FW, and if the next release is improved as SansaFix indicates it will be, that is even better. Way Cool.