That's where the speculation gets interesting, and a lot of different constraints start interacting with each other to determine what comes to market. I look at my Apple Watch, which uses about a third of it's battery through a normal day with ~18 hours on my wrist, and figure that if you dumped the optical HR sensor and cell modem along with their power draw and gave it a battery the same volume as a 530/830, it'd have no problem matching (probably exceeding) the "up to 20 hours" Garmin claims. I think that Garmin (and Wahoo) are largely constrained by the simple economics of how many people out there are just the right level of "serious" who both want a cycling computer for the data, navigation, or both, and decide not to use their phone (which are now efficient enough to not eat their batteries with a few hours of GPS use).
I don't expect the same level of hardware or design sophistication from Garmin as I do from Apple, I do expect reliability. And Garmin's software is 90% of where they fall down in that department. I don't know if they need to hire more people, or get better people, or change their development process, or culture/management, but they've gotta do something to make it better.