Lasted under a day of light usage.  Very unimpressed.  Yet another reason to tear it open and see what's going on!  Also after looking at the collar connections again it appears there's only three signals, presumably power, ground, and some form of one-wire protocol to read/write the EEPROM.  That rules out the potential match for the EEPROM in the last post.

The battery LED blinks twice when the cartridge is running low (according to what they consider low, at least) and will signal when it's 'empty' as well, so there's definitely some smarts in this thing, probably a small MCU in the battery that reads the EEPROM in the cartridge, handles the LED functions, and then passes voltage to the atomizer if the value on the cartridge EEPROM is greater than zero..then of course deincrements the value.