I've used Globalstar's "Satellite Data", Inmarsat's BGAN, and Starlink offerings for years (both commercial and in public safety), and it's difficult to put in words how terrible & deficient Globalstar's offering is in terms of the kind of traffic that "5G Satellite Internet" embodies.
To be fair, Globalstar's network was never really meant nor designed for that. It's meant for store-and-forward kind of async, small message comms, not TCP/IP traffic. Everything from the sats, the orbits, the constellation density, the terminals, the spectrum, the protocols they use has design decisions to favor asset tracking, very short message, etc with incredibly low power use (idle & peak). The devices are incredibly rugged, cost effective for their profile, and reliant on field-servicable kit, too.
OTOH, Inmarsat's latency is high, but bandwidth is good (if you can afford it), but it's sats are so far out compared to LEO constellations, but for circuit-switched data (video, audio, etc) is actually pretty good. But capacity is pretty damn low.
Starlink's constellation, whether you like Musk or not (and SpaceX/Starlink has thousands of employees that do the *actual* work how are *NOT* Musk and I know many of them who don't agree with or believe in his public antics and are still good, smart, wonderful, nice people), is pretty much ideally designed for this kind of stuff.
Bottom line: anyone thinking Globalstar can pull this off without a major change in their entire network either knows something the rest of us don't or is insane.
