StarLink provides a great oportunity for politicians to delay or cancel projects which would otherwise have given broadband connection to underserved areas. In urban planning this is known as the Uber effect.
Depending on the area of the world, wireless and other options that exist that are likely sub par. It is on every continent including North America.
Some regions of the world have aggressively invested in fibre in rural areas.
We can see in parts of the world where there is a lot of investment (and interactions with govt for permits, etc) in physical infrastructure, whether its coax for cable tv and internet, copper for phone lines (and ADSL), wireless doesn't always have a nice way in.
There are places in the world that didn't get as much wired infrastructure put in and were able to jump up to much better wireless.
Satellite based internet as a category provides an additional coverage where "traditional" infrastructure hasn't made it yet. This can be wires or other wireless.
Broadband internet via cables, fiber optics, and radio towers is state of the art in telecommunication infrastructure. Satellite is both slower, more limited, and more prone to various disruptions. The capabilities of the wires and the radio towers is also improving. 5 years ago we didn’t have 5G towers, and 20 years ago fiber optics seemed a distant dream. The only thing freezing traditional telecommunication infrastructure in place are dreams of low earth orbit satellites which will never materialize.
If I understand your analogy correctly (which I‘m not sure I do) this is like looking at the new technology of pneumatic tubes and stipulating that all postal delivery will be done using this new technology in the future, and we may as well stop funding the national postal service, remove mail-rooms from our ships and trains, because somebody will build a pneumatic tube that will deliver mail door to door between New York and Chicago.
Do you truly believe this statement, literally?
When I say LEO satellite internet won‘t materialize, I mean that it won‘t serve everybody. It will always be an expensive option which in best case will be subsidized for only a portion of the people that actually need it. LEO satellite is not the future of telecommunication infrastructure, it is lacking in almost every way next to traditional infrastructure. The only thing it is better at is a) marketing and b) providing internet to rich people in their yachts or in their mansions 10 miles out of the suburb.
If SpaceX sells Starlink to Amazon and buys Corning, Ciena and Nokia, they'll be extolling the virtues of LEO Megaconstellations
