a button adds weight (just a little), takes up space leading to a bigger design, adds unnecessary complexity, and any extra components add another place for something to break. Also buttons fail, a failing button could press itself, and reset the tag, meaning it's no longer tracking what you thought it was.
it would also add a very easy and obvious way for a "thief" to disable a tracker, making the whole thing useless. Although if a thief had found the AirTag and could press the button, they could just as easily discard the tag.
I've had 3 AirTags in use since they came out, and I've never had to reset one.
If you took a poll and asked how many people have had to reset an AirTag, the number would be quite low, especially if you compare it to the time of use. for me it's 3 airtags for the 230 or so days since they came out, so I'm around 700 days of use without a reset. if you were to take all the tags in use, and the total number of resets done, that number might be a reset every 10,000 days of use. Why increase weight, size, complexity and points of failure for something that's used once every 30 years of use? (at this point that would be one reset for every 45 AirTags out there)
there is a way to reset them, it's easy enough to do if you need to, and hard enough to do that you won't accidentally do it.
I can't repair this airtag,
do you have an advanced degree in electrical engineering? of course you can't repair it.
can you fix your TV remote, the clicker for your car, or your cell phone?
it's a tiny fairly complicated device, that designed to get a decent amount of physical abuse (dropping your keys, being sat on), that reliably runs an rf transmitter for over a year on a coin cell battery
although for an AirTag, doubtful even an EE could repair it, they'd probably be able to diagnose what went wrong with it, but trying to remove components would be near impossible, and would more than likely cause more damage in the process
