You also have to collect almost every droplet, prevent the spray from drifting, and keep droplets from blocking one another's radiation. These are significant engineering challenges. Which they already use on starlink.
Until all deserts, rooftops and roads are covered with solar panels there is zero case to move things to orbit for freer energy.
Also data center cooling consumes around 40% of power, not 60%, and in cooler climates you can do much better than that.
Claude Shannon would disagree.
With fiber you can cheaply add capacity by upgrading terminal hardware where needed. With Starlink you can only add capacity to the whole world at once, and it's mostly seas and woods.
It would appear that every commercial ship and every RV has a starlink already, this is perfect application for starlink. Competing with city ISPs would require thousands times the bandwidth at lower prices, it's a non-starter.
I think the more interesting comparison is marginal cost. Fiber is extremely scalable once installed, but passing an additional household can be expensive. Starlink avoids that last mile construction entirely.
Explain to me how that can satisfy say Paris.
What's your business / engineering case for Starlink (i.e. LEO satellites) dominating the internet market worldwide?
I was an early believer in LEO internet delivery and realise its advantage over terrestrial delivery in some regards, but I haven't seen reason to project its 'dominance'. So I'd be interested to know why you are projecting this.