I am working on a new meter that measures voltage (a simple static guide), Amp/hours consumed since recharge (aka coulomb-counting) and has an audible feedback on the actual current being consumed.
Coulomb-counting is a good measure, but you need to calbrate it as the battery tails off over charge cycles. That can be done dynamically by storing how many Amp/hours came out between maximum voltage (when freshly charged) and when the controller cuts out (usually 30V or 20V depending on battery nominal voltage)
The 'proper' way to measure charge state is to measure its impedance - how its terminal voltage changes with load, and so the intuitive way is to ride trying to keep out of the Amber!
The audible feedback (akin to the ticking of a freewheel) is intended to help the rider manage the power without looking away from the road! The voltage measurement gives some idea of charge state without actually sucking power (SLA, NiMh and LiIon all exhibit falling voltage with charge.
I remember my Father had a vacuum gauge on the inlet manifold of his Hillman car and drove to achieve least depression - I had a Jag. with an 'econometer' that did just the same, but why would you want that on a 3.2S!!!
I am a reformed character now....