Batteries naturally lose charge when shelved or when disconnected from their power source. It's just physics, they decay. I argue that the battery life measuring logic is different from platform to platform, mac to PC and PC to PC which is the root of so many messages I've seen pop up lately.
The SB has batteries all over the place, they obviously have more logic in the management, measurement and care and feeding of the lipo cells in the SB than in many of these other devices that only have one.
Even further, if that's not the case, I'm willing to give MS a pass on loosing 1-2% over (say) 15 hours of fully powered off state. Even Hibernate state. This is a rounding error and there's so many variables to consider. I.e., was the system operating on battery at the time of measurement or A/C? Were the cells in the system well cycled or always powered to 100%. What temperature were they at? It all matters. Surprisingly well cycled lipo batteries have a longer life than ones that spend their lives fully charged, and they live longer in the middle of their charge cycle, not at holding 100% for longer. So these 1 and 2% issues are just noise.
The problem with these last couple days of messages is that it's distracting from the issue. Standby is broken, absolutely does not work. Hibernate is a work-around to the defect, not a solution. And what can we do to get MS to put more effort into fixing it. Money talks with any corporation. I will be returning mine at the end of my return period and most likely ordering a new one. For each return there is much cost of re-inventorying and refurbishing and reselling. If you want to help Microsoft fix the problem, return your device and let them see increments on their defective SB returns. I love the platform they built, but we all agree they need to fix some pretty large new product issues before it's one I'll take the deep investment hit to own.