Thank You again. It seems that i must check min resistance limit on my own - but thats no problem. Now its time for little waiting - I'm going to perform further tests, when my ordered equipment arrive. Want to be 100% sure, that I'm not going to slowly destroy anyone battery by suggesting new method...

Early test were successful - was able to hotswap battery in field without any thing connected to USB port (in fact, second test proved that hotswap is possible also during hostmode), without disabling BME, with WIFI enabled downloading file at 10 Mbit/s + cellular 3G connection enabled and browsing pages, with brightness set to full. So, as i presume, there won't be need to disable anything in real life situation. Stay tuned.

so any progress? (sorry for resurrecting this but I was curious)

Hotswapping with very short (for sure under 0,25 second, via mechanical switch) paralleling of main battery with other small li-po - both behind amperage limiters - works great. Practically, it is 100% safe, due to very short time of paralleling + limiters = no way of extensive energy transfer between batteries (it's like one battery is charging another for <0.25 second...), but theoretically, it can be dangerous, so I don't advocate this approach.

In practice, I'm using mechanical hostwapping mechanism (inside Mugen Cover, no external parts) for > month on almost every battery discharge, and I don't experience any battery issues (voltage, internal resistance, capacity - all in norm). At the same time, i'm experimenting with capacitor design, that eliminate paralleling all-together. the problem is - most of the time - with size of capacitors Especially, that we wan't hotswap no matter of load. Seems quite impossible to pack up enough capacitors inside Mugen Cover, at least for reasonable price. Probably will change in future, ho ever, after tests I'm quite sure mechanical way is safe. Still testing and observing, though.

/Estel