Questions that should be asked

I think it is going to be more efficient to process the local video and motion sensor data locally than to ship the live data to the ground and back again, at high latencies, just to maneuver a floating camera platform.

Also according to the article AA batteries are used because they are standard on the station, so this is a resource they already have, not something new just for this device.

Even if these phone devices only last a couple of years they are probably still worth it, they aren't very heavy to ship and the cost/benefit is likely not worth spending a ton of engineering time and design money to build something more robust just to save on the cost of 1-2 pounds over the course of 10 years or so. I'm sure there are other wear items which are heavier and have a higher engineering priority.

Over time I'm sure the design of these kinds of devices will improve, maybe they'll start buying their own SoCs and building their own cases but that better future isn't a good reason to deny the good enough solution that provides benefits now.