There are a lot of experimental batteries that have like 5x capacity of the current Li-ion ones, but melt and explode when heated or crushed, which makes them effectively useless.
This is one of the hardest problems in high-capacity batteries - to avoid "rapid unscheduled heat dissipation" and the following "rapid unscheduled disassembly".
Animals can store a lot of energy, but they don't tend to explode with that energy when damaged. Are there techniques that could be learned from cells that could be used to make safer electrical batteries?
Then there is Li-Po which can ignite if charged too much or if left to drain to low. I also suspect they'll burn if you give them a stern look or make a hurtful comment.
As you have it isolated from the house in a specially built enclosure or garden shed it's less likely to get damaged and even if it is the damage can be manageable.
Then again in these cases it's more of a matter of how cheap it is vs the amount of energy it can hold and size doesn't matter a much.
1: http://goodforgas.com/hazardous-gases-associated-lead-acid-b...
Say solar, kinetic, or wireless charging stations spread throughout a city? Maybe incorporate some kind of "witricity" into the next WiFi proposal? If stuff can be recharged while it's being used, or within seconds, it shouldn't matter how much capacity the battery has.
But that’s exactly what Li-ion batteries already do. x)
http://www.candlepowerforums.com/vb/showthread.php?120888-RO...
TL;DR: guy used stock flashlight with two stock CR123 batteries in series. One discharged faster than the other, so it got reverse charged from the one holding more charge, and promptly vented hydrogen gas that got ignited and blew up. Guy got glass and metal shrapnel damage to his foot, and hydrogen fluoride poisoning resulting in permanent lung damage and other bad symptoms. It's really horrible. Bad stuff starts at page 5.
With nasty cough as a bonus