Regarding "only needs to be solved ... high density" - Musk has gone on record
as saying that Starlink probably will never support dense subscriber areas.
Which of course contradicts many other statements by Starlink and Starfans that
they can scale up to full coverage of the world. My point
I am very wary of a generalization of this problem: software engineers who
believe that they can code around arbitrary idosynchronies of network
hardware. They often succeed, but generally at a severe performance
penalty.
How much do we know about the actual hardware? As far as I understand