They're also certainly not providing internet to passengers via Iridium, because Iridium data is extremely, hilariously expensive. Sending back basic airplane telemetry via data satellite is quite practical though because it's a very small data volume, and is already often done by engine leasers (like Rolls Royce and GE) for continuous engine performance monitoring. It's rather depressing that on many aircraft in the air today, the engines are better monitored than the actual plane.
There are different motivations though. It pays for GE and RR to collect constant engine telemetry so that they can micro-optimize service intervals (read: perform service as infrequently as possible). The airlines, though, have physical possession of the planes and tend to figure they'll work that kind of thing out on the ground when planes are out of service for several hours anyway.
The price of launching a set of satellites that covers that area divided by the number of customers that will use it over the number of years that's in service? That's a price.
One issue is getting the different orbital planes covered. Its possible with an "upper stage" booster[0], but that is significantly more expensive than a typical CubeSat carrier/deployer[1] for the ride-along launches. As far as I know, there is no current smallsat/cubesat sized upper stage intended for changing orbital planes.