Fiber beats starlink by orders of magnitude.
With coax it depends on how it's set up. Plenty of existing lines have worse upload speed than starlink because of frequency allocations that are expensive to change.
Phone wires are very bad at carrying data and are usually much slower than starlink.
That's just in U.S. Starlink is world-wide.
Back of envelope calcs:
$100/month == $1k a year.
1 million * $1k = $1 billion
10 million = $10 billion.
Estimated cost to replace all Starlink satellites every 5 years: $10 billion.
That's $2 billion a year.
Cost: $2 billion/year. Revenue: $10 billion/year. Profit: $8 billion/year.
Assuming only 25% of Americans currently without high-speed internet will opt for Starlink, not counting world-wide users, not counting commercial applications (boats, planes, gas stations), not counting using Starlink as internet backbone (i.e. alternative to under sea fiber).
Most likely Starlink will be a cash cow (and finance Mars colonization).
https://www.npr.org/2022/05/11/1098368187/42-million-america...
42M Americans = ~10M households (assuming 4 per household).
Starlink is $110/month and upfront $599, which puts it outside many/most of these remote folks.
10M household is the TAM, so SOM would be just 5-10% of that, about 500K-1M households, or even less.
Regarding the cost of replacement of satellites and their operational costs, I have no clue.
In other words, how many people with hard wired fiber/coax will change to Starlink or will be be useful only for the rural/remote segment.
Yes, there's some international, marine, flight etc. market, but, again, how big is that.
Serious question - is there any real data?