In the interest of completeness, and since I still see an Apple Magic Trackpad shown on the Batteriser website as a device that stands to benefit, I am providing the testing data for this as well. The test setup is identical to that of the Apple Wireless Keyboard I posted earlier. The short answer is that the minimum device operating voltage is 0.95V per cell, and the cutout voltage is 0.925V per cell. The same caveats of voltage drop at the terminals as before apply. In short, the currents and ESR involved are low enough to be largely irrelevant to the result.
In the end, the Magic Trackpad is even more efficient a battery consumer than the Apple Wireless Keyboard and leaves less than 5% of the battery's energy remaining when the device is run to the point it ceases function.
V % Notes
============
3.2 100
3.1 100
3.0 100
2.9 93
2.8 84
2.7 76
2.6 58
2.5 42
2.4 ----- no measurement taken
2.3 17
2.2 13
2.1 0
2.0 0
1.9 0 min startup voltage
1.85 0 Device ceases function
At this point, I feel that Batteroo and it's advisors, particularly one Dr. Kiumars Parvin, Professor of Physics at San Jose State University, need to demonstrate publicly how they have come to the conclusion that these devices leave 80% of the battery's energy left untapped, and that a boost converter can release it. I am using what I feel are reasonable methods that are backed by standard calculations, and I am completely unable to reproduce Batteroo's results.
My offer to reproduce these tests before an audience of qualified observers stands. You come to me. I only ask for reasonable time accommodations.
====================================
BTW, Dr. Parvin has a impressive CV and has authored nearly two dozen peer-reviewed publications:
http://www.sjsu.edu/people/kiumars.parvin/http://www.sjsu.edu/people/kiumars.parvin/CURRICULUM%20VITAE.pdf