SpaceX Starlink Constellation, page-27

There is a good summary of the SpaceX Starlink in the SGU's Feb 24th podcast. Sorry their site is playing up ATM so I can't post a link. Fast forward to about the 13min 40sec mark to hear just the SpaceX part (goes for about 8 mins).
Some of my takeaways from it:
- Their constellation is planned to be operational next year. (Seems ambitious given they also plan 8 months of testing first, but I'm not sure what exactly they mean by "operational")
- The target market is initially the 15 Million Americans currently without access to broadband internet.
- Also will be targeting up to 10% of internet traffic in US cities to ease internet congestion.
- Planning 6 Mobile ground stations.
- Alphabet (ie. Google) have invested US$1Billion to the project.
- Elon has said this could provide internet to undeveloped areas for the first time.

My opinions is that whilst this constellation will provide the availability of broadband globally, the target (at least initially) is the developed world. Apparently 15 Million people in the US are without access to broadband. I know areas of Australia are without adequate broadband services and I can only imagine parts of Europe, Canada and developed Asia and Latin America would too have similar issues.
A threat to SAS's is SpaceX could practically give the service away for nothing in the developing world as an act of goodwill or just to recoup a little more money that would otherwise go begging.
With the math I've done I don't see Starlink making money or breaking even for that matter in the short to medium term. Maybe SpaceX is playing the long game and see the returns a long way down the road as a dominant global ISP, maybe investors like Google (and possibly Facebook) see this as a good long term investment as it could get billions of extra users onto their platforms.
It's important to note that SAS's entire constellation will cost about the same as a single Startlink Satellite. So while the consumer side of the market could be disrupted by a superior broadband service (if affordable) the low cost of SAS's service will still be attractive for niche industry applications that don't require broadband like SCADA, M2M, IOT and tracking.