This may seem off-topic but it is not:
Earlier today I was reminded that until early 1998 I was severely hampered by total lack of communication from an island in the Bahamas I then owned. My work at the time demanded constant telephone contact,so I was often forced to take a half hour boat ride to the nearest large island then drive another half hour to get to a telephone. Then an acquaintance who worked with Hughes told me about a new service called Iridium. It had a ~10 pound base station, a sizable antenna and an instrument not unlike the Motorola DynaTac I had been using around ~1985. I bought it and subscribed. The hardware cost thousands and the subscription was hundreds per month plus a few dollars a minute. The lags were horrendous but I had a telephone. Then a year later while I was off-island and my spouse there, Hurricane Floyd hit our island as a strong Category 4. We had power thanks to our solar system, and iridium worked too, so my spouse and I did maintain contact.
Now think of how many places in the world there are which are out off reach for traditional communication services plus ships and airplanes. These days many of those have some form of Satellite communication available but often at very high cost and high latency. Now comes Starlink.
Nearly all the discussions about Starlink emphasize better service where existing ones are deficient, expensive and unreliable. I think the most profitable single category for Starlink will be those. The second one will be emergency services (e.g. Washington State) plus military and governmental communication. Only third will be the one we all think about, our homes and better internet plus perhaps our cars.
All this seems likely to be 3-5 years out for anything remotely like global coverage, with the majority of that time dealing with regulatory impediments and point-of-use sales/service/feature issues occupying most of the time. Soon it will be time to try to quantify this potential.
I suspect Elon thinks of going public once the bulk of all this is appearing and any foolish analyst could see it. This market will be much like Autobidder, and not unlike the growth of near global conversion to BEV. Competition is already emerging but not very quickly and not yet scaling. The biggest single competitor might just be Amazon.