With Starlink alone SpaceX has crazy potential. They're severely underplaying customer base potential in order to avoid massive lobbying psuhback from traditional Telcos. In five years from now it will be clear that SpaceX is an existential threat to even the largest of providers, even if they can't capture NYC/LA.
Agreed that Starlink will be a significant player in providing internet service (or the pipe for internet service), but one cannot dismiss terrestrial service providers as the competition for users in traditionally low density regions heats up. Its not like Tesla vs the auto industry, where the established players have basically waited for Tesla to lay the EV groundwork while they leaned back on brand recognition/loyalty. There's basically no such thing in internet service--nobody cares where it comes from (other than maybe identifying the one place they
don't want to get their service from is their current provider) and the connectivity establishment understands their product needs to keep up with cost-performance in a way that the auto stalwarts never could.
In any case, here are the potential threats I see to Starlink's business model:
1. There will
always be a threshold of user density at which terrestrial providers can build out infrastructure to sell better+cheaper service than Starlink. Plus as both population and density increase over time and terrestrial providers find it profitable to expand further into traditionally rural areas, the potential starlink user base will continually shrink. Even the initially beneficial distribution of WFH workforces will eventually turn negative on Starlink.
2. There will always be a global entry threshold for users that's a function of wealth (can someone actually afford internet service?) and politics (will a country actually permit service in their borders?). I expanded on this over in the starlink thread, but the TLDR is that there's a significant global population that doesn't qualify, and in the end a majority of the global Starlink users will be in the USA.
3. Next gen/mm wireless networks: Obviously there's plenty uncertainties with and plenty of work to do on that technology, but the bottom line is a similar story around user density. If a terrestrial service can last-mile wireless from a few towers that's going to be a game changer for providers that find hardwiring users too expensive. This has an additional threat if direct-to-device service turns into a reliable thing.
4. Perhaps counterintuitively, there's no current competition in space. OneWeb is restructuring and the tech was always going to be lagging Starlink, Telesat is years behind assuming they even remain a thing, and the establishment GEO internet services rely on technology that even in latest-and-greatest form (which is still conservative) is ~twice the capital of Starlink. (I did a hand-wavey cost normalization of the two in the starlink thread if you're interested). BUT, as the history of anything tells us, no advantage lasts forever. As more and more companies try to compete in the satellite internet space, each company is going to have to settle for a smaller piece of the pie and/or evolve their product in some differentiating way.
Just spitballing a potential disruptor to Starlink's otherwise massive disruption, many folks see GEO as 'dead'...but IMHO its actually ripe for opportunity. IMHO the above mentioned latest-and-greatest GEO internet satellite only normalizes to ~twice as costly as Starlink because nobody like SpaceX has ever tried to do GEO like SpaceX has done LEO. As a data point, SpaceX has figured out a way to reduce cost of a LEO satellite by over 2 orders of magnitude while also
increasing capacity by more than 3 orders of magnitude (using Iridium Next...kind of the only relevant comparison that's in public domain). Obviously a big part of that comes from economies of scale, but if a GEO (or MEO) constellation could realize even a fraction of those gains on a new school vs old school GEO/MEO, that math very much checks out. Partner with a LEO network (OneWeb or someone new?) to handle the fraction of traffic that actually needs low latency and its a least plausible competition to Starlink.
That's not to say I think Starlink is doomed--far from it. Folks who know me know I'm very bullish on Elon's unparalleled genius and vision, and I firmly believe he's not going heavy into something that doesn't have a significant probability of success. I just happen to also be a little more realistic on how 'success' is defined. It's obvious that over the next few years as Starlink builds out their constellation they're going to have some pretty blue skies; In that same time, we'll learn which if any of the above threats turn into thunderheads...