It seems like the thesis is if cost per kg fall enough then the implausible becomes plausible. F9 reduced costs to make starlink possible, what will starship make possible?
Seems to me as if Starlink without Starship isn't all that great.
Also, OneWeb - despite being smaller in scope - managed to do a similar thing without reusable rockets. Now that doesn't make F9 bad - it's still a great machine - it's just not the game changer many hoped it to be.
It's about 1,500 active satellites versus 600 + ~50 spares - quite comparable.
It's pointless to base your assessment on plans if the past showed how quickly they evaporate in light of new developments.
[0] http://spaceref.com/mars/spacex-will-start-going-to-mars-in-...
[1] https://www.nationalgeographic.com/science/article/spacex-mo...
There is a reason they went bankrupt and were mostly solved by government bail-outs. Because its in fact not viable as a business if you have to launch on traditional rockets and have contractor produce your sats.