According to ITU, more than a third of world population has no (proper) access to the Internet, and that equates to about 2.7 Billion people worldwide [1].
Personally I'd really love Starlink to provide Internet access to all people in the rural areas but it probably will not going to happen. By their own estimate, they are only currently deploying around 5% to 10% of their 42K projected Starlink satellites for global coverage. I'm not seeing that going to happen for at least another ten years (2033) provided they get the funding they're requesting and it does not seems to be the case [2].
The best alternative for satellite link for rural is the terrestrial based fixed wireless access (FWA). There are mainly 3 segments/types of FWA namely [3]:
1) ‘Wireless Fiber’ (100 – 1K Mbps, cost USD50 – 100)
2) ‘Build with Precision’ (50 – 200 Mbps, cost USD20 – 60)
3) ‘Connect the Unconnected’ (10–100 Mbps, cost USD10 – 20)
This last (3rd segment) is characterized by virtually non-existent fixed broadband alternatives, and mobile broadband using smartphones is the dominant way of accessing the Internet. Average revenue per unit (ARPU) cost are very much limited and user expectations of access speed are relatively low. The irony is that the 3rd segment is the "low hanging fruits" in term of technology while being the hardest to achieve in practice due to the very nature of terrestrial wireless signals. Starlink or any LEO based solution, however attractive, cannot possibly fill this gap due to the high cost involved in the satellite business and its astronomical ROI.
The author of the OP article is from the first (1st) segment of "elite" users and they're very far in between in the grand scale of rural users, i.e. not a typical rural users most probably 5% maximum. It seems to me that Starlink is focusing on this type of "elite" rural to survive and remain profitable not unlike Tesla did with their initial luxury/premium Tesla EV models strategy. But Tesla success is largely hinged and depended on huge (US) govt support for Tesla early days, and Starlink is also counting in the same strategy [4]. However, the latest bidding for govt money for Starlink is not successful and it's evidence from the stalls number of further starlink satellite deployment [5]. Elon Musk has been mentioning on record that one of the first Starlink priorities is not to go bust as what had happenned to its predecessor Motorola's Iridium but the odd seems to be against them at least for now.
I think it's the best for most of the countries that struggling with rural internet access (e.g. all countries US included) is to go for terrestrial fixed wireless access (FWA) for the third (3rd) segment i.e. connecting the unconnected. But if US cannot make satellite based Internet access work what are the chances that other countries might have?
Probably the best solution is to be flexible in tackling the Internet access for the 3rd segment with FWA perhaps similar to approach taken by Magma [5]. It's still early days for the rural areas connectivity but the sooner we can tackle this issues is better, as demonstrated during the height of the pandemic where travels are limited and digital divide phenomenon is at its worst.
[1] Measuring digital development Facts and Figures 2022:
https://www.itu.int/itu-d/reports/statistics/facts-figures-2...
[2] FCC decides against giving Starlink $1b in rural broadband subsidies:
https://www.theregister.com/2022/08/11/fcc_starlink_rural_in...
[3] Fixed Wireless Access handbook (2023 Edition) by Ericsson:
https://www.ericsson.com/en/fixed-wireless-access
[4] SpaceX’s Starlink satellite network wins $885M in federal aid for rural broadband:
https://www.geekwire.com/2020/spacexs-starlink-satellite-net...
[5] Building Flexible, Low-Cost Wireless Access Networks With Magma (to appear in 20th USENIX Symposium on Networked Systems Design and Implementation 2023):
https://www.cs.princeton.edu/~jrex/papers/magma23.pdf