Sure, there is a tipping point, but that point may never (can never) be reached in many areas. Very few are going to pay multi-thousands to get cable/ fiber vs Starlink. Nor do many areas allow the population density to cause the companies to install it on their own.
No disagreement. There are always going to be edge cases where Starlink is the only, let alone best option. There are always going to be corner cases like the random Boston article you linked. There will
always to be a user base for Starlink. The point (in this case) is that its foolish to assume terrestrial providers are going to sit back and let Starlink scoop up all the rural users without any fight.
Starlink being available before terrestrial service means it will be harder for terrestrial to expand (if it even wanted to).
Why harder? Agree there needs to be a desire from terrestrial to expand (where desire = profit), but its not like there's any material first-to-market advantage for starlink. Broken record here, but people don't care where their service comes from.
Hell, if anything an operational Starlink network gives the terrestrial providers high fidelity targets and a good sense of user distributions to input into their cost models to determine if/when its worth expanding. It will be especially interesting to see what happens in the space where people are dropping their shitty terrestrial service for Starlink.
Starlink can support this also, without a hard line being run to the tower.
Of course--we've talked about this many time before in this forum. And at an LTE equivalent level of service the starlink backhaul is almost certainly going to be more than sufficient. The question really revolves around the number of users a starlink backhaul can support at higher levels service...let's say the standard-is 100/10 metric. A couple people? No problem...though its pretty silly to build a tower for that. Ten? Sure. Many tens? Certainly not with the initial constellation phases, as the that effectively creates the user density problem that would suck too much from neighboring users. For this one the good news is that as the constellation grows so will its capacity, and at a rate that's likely greater than the sum of its parts (in this case, the sum of the individual satellite performances) so it would be easier to aggregate many tens-of-users towers across more sats. But again, the real question is at what point it makes more sense to hardwire the tower vs strap a starlink terminal on the top.
GEO/ MEO has additional issues. Starlink has savings specifically because it is close:
Assured deorbit
Less transmit power/ receiver sensitivity (both ends)
Tighter spot for same size antenna. Along with this, higher levels of bandwidth reuse
Significantly lowing ping times
Lower launch cost/ more sats per launch
Sure, and I'm a LEO person (formerly a long time GEO person) so I'm not actually making a hard sell on GEO or anything. I'm just trying to be fair in assessing the potential for GEO as opposed to the current reality of GEO. Again, nobody like SpaceX has ever tried to do GEO/MEO, so its not quite a slam dunk to say there's
no future up there. I mean, everybody thought EVs were crap until Tesla came around. Everybody thought rockets were expensive until SpaceX came around.
I'm also not going to dismiss the benefits of GEO/MEO:
-Practically no deorbit (needs barely some leftover ∆ v just to raise up to a graveyard)
- larger satellites = lager solar arrays, + shorter eclipses = way more available power
-WAY bigger antennas = higher gain (to offset ranging losses)
-Larger coverage area = way fewer satellites
-Fewer satellites = all manner of efficiencies (think: moving cargo with a train as opposed to a bunch of tractor trailers, think: we only need a couple reaction wheels instead of thousands)
-Basic UTs = inexpensive UTs (no need for phased arrays that have to track wide swaths)
-Maximized operational duty cycle (no 'wasted' satellite capability while they're flying over low/no user areas)