Is depletion of the glucose oxidase reagent the limiting factor, versus the ~~17 week estimated battery life? One would expect an item with a 17 week battery that also has the glucose oxidase reagent electrode that also lasts ~~17 weeks to have a higher service life - since it does not, I expect the electrode is the limiter and it might have perhaps 30 days? Of course, it has a clock and a limited life to guarantee correct service beyond the 15 day clockout. It will take a better hacker than me to amend the software to explore these electrode limits in a safe manner. One would expect a foreign body might face attack by the body that inexorably acts to cover/passivate this foreign body.