This WILL work, but only for circuits with extremely low drain and batteries with very, very low self-discharge rates (not NiMH batteries, for example, which have pretty high self-discharge).
Basically, let’s suppose that a Bluetooth keyboard (on average) draws 1mw at 3.0 volts and won’t work below 2.6.
When an ordinary battery crosses the 1.3v/cell threshold, it might still hold a small, but significant amount of energy that you wind up throwing away.
What they’ve done that is fairly novel, is to make a tiny, very efficient, very, very low voltage to very low voltage DC-DC boost converter. I wonder how low on voltage/current they can go to get 1.5 volts, and at what efficiency.
This WILL work, but only for circuits with extremely low drain and batteries with very, very low self-discharge rates (not NiMH batteries, for example, which have pretty high self-discharge).
...
The better NIMH batteries are now low self-discharge, and NIMH batteries have a flat discharge rate, so this device wouldn’t improve them. The same goes for lithium batteries. So what this device may do is make alkalines discharge more like NIMH and lithium batteries.
Some LED flashlights have had similar boost circuitry in them for years.
Likely an oscillator & multi-tap transformer circuit. The concept is used for single battery LED flashlights, high voltage generators used to charge capacitors for EMP simulators. and battery operated tube radios.