Nearly half of humanity doesn't have Internet access, not a few rednecks in the boonies. That's billions of people, the world is much larger then the US.

And it's not even just some rednecks in the boonies in the US either. My dad's house in Maine is just 10 minutes from a fairly sizable city (by Maine standards). He can get maybe 1Mbps down and the houses further down the road can't get wired Internet at all. Cell phone reception is also almost non-existent.

Being ten minutes from a city, it seems a single cell tower would solve the issue much better than a satellite no? In that instance, satellites feel like a work around for poor development practices (whether that be city or state regulation, or provider investment) than a reasonable solution.

Maybe there is not enough houses in that area to spread the cost of that one wireless tech hosting tower?

A single satellite from a swarm will cover much bigger area at any one time, not to mention the coverage area moves as it orbits.

Much better argument for paying infrastructure costs if coverage area is effectively "humankind".

Maybe, there aren't enough houses to justify the cost, sure. I figured that probably wasn't the case given the proximity to a city.

I'm not arguing that a satellite solves the problem of remote data access generally, they definitely do. My original question was whether or not that specific use case is worth the cost to science.

They don't solve access generally, so the other threads which reference large swaths of the planet not having internet yet feel beside the point to me. The vast majority of humanity lives in densely populated areas, and therefore these satellite based systems still won't solve that problem.

Yes, but these constellations aren't designed to serve 'everyone without internet', they are designed to serve areas with low population density where physical connections are financially infeasible, hence my usage of 'boonies'.

It's perhaps worth noting that Starlink is about much more than network access in remote areas. The laser links between satalites should enable a considerable improvement in intercontinental transmission, both bandwidth and latency.

That's simply not true. The throughput provided by starlink absolutely pales in comparison to an optical fibre link. AFAIK current plans might look at 100 Gbps transponders for the interlinks (I think it might even just be 10 Gbps). Even if they go to future 800 G or 1T transponders (which is questionable if it's possible, especially because most of the constellation needs this, so a whole exchange of satellites would be required), that's still much less than 70 Tbps that can be transmitted through a single fibre over submarine distances, and a submarine cable typically contains tens to hundreds of fibres.

So in short starlink will have no influence on available crosscontinental bandwidth.

Source: I research on optical fibre and space communication

If you're using the satellites purely as continental links, then the overall number needed would be much much lower, as you could plan service areas as opposed to trying to services the entire surface of the planet.