Awesome to hear about these new developments! I can't wait to start playing with them. If I may make a suggestion here... an exponential back off might be an interesting idea to explore. Halve frequency of broadcasts after some amount of time, then halve them again after a block of time, then halve them again... etc. Maybe stop when you get out to some ridiculously long time between squaking... possibly make it user configurable what the slowest update frequency once the rocket is at rest will be? I know if I'm the only one at the club flying with your stuff I'd want a longer time than if there were dozens of others that could possibly see it while in the air. The longer it's been, the less likely that someone will be coming over the horizon *right now* looking for it. I'd even suggest powering down GPS once the tracker is "at rest" to save battery for broadcasting.
Oops, I forgot to replay to this a while ago. After a flight and landing, if the tracker doesn't hear from the ground station after 5 minutes, the tracker goes into lost rocket mode. In this mode, both the GPS and the main microcontroller go into a very low-power sleep mode in between transmissions, and the length of each sleep mode does increase as time goes on. The Bluetooth microcontroller, however, stays awake, so unfortunately improvements in battery life are somewhat limited. But as it is, a standard 400 mAhr single cell will keep the tracker running through the rest of the flight day. Later if I can get the Bluetooth microcontroller to also go to sleep then we should be able to extend the life of a 400 mAhr battery to days.