And, anywhere with trees around (e.g., here), it doesn't work.
And, anywhere with trees around (e.g., here), it doesn't work.
On the “financially sustainable” question: Current public reported runrate is $40+mm/month in revenue = $500mm/yr ARR business on roughly 400,000 customers. Customers are still limited by geo for the reasons you mention - they need more satellite density, so I presume more customers would sign up than have right now, e.g. there is a lot of growth ahead in raw revenues. That should give the current business, as is, at least a $5bn valuation, frankly quite a lot more given the ease with which they can sign up whole countries worth of people.
Can starlink add another 400,000 customers for less than $5bn? Absolutely. SpaceX has paid the capex for lift already, and right now it’s a matter of adding satellites and creating antennas and paying (at-cost) lift rates for the cheapest lift known to man. Starlink is going to be a $100bn+ business IMO very quickly, like next 5 years.
Per your point, e.g. ‘is it good enough near trees’, or what percent of the earth can Starlink serve? On the one hand, it doesn’t matter a lot because that number is clearly north of 10%, and that makes the starlink broadband market 3 times the size of the US as is — and that ignores the various super-high-value markets like shipping, jets, and elsewhere that will pay 50x a ‘typical’ customer - but, I think the tree/coverage problem in general is a problem of low satellite density right now.
My starlink is laid out as best I can near my house, and it’s at like “96% coverage” per the app; every ten minutes or so, it loses connectivity for a bit. That’s because all available satellites are occluded by my trees - with additional satellites, I wouldn’t have that problem. So, I would guess that problem gets less bad over time as they grow.
And by my estimate they need a million subscribers just to pay for replacement satellites as the old ones burn up.