Maybe, but my sense as an outsider was that Starlink was moving at a snail's pace and was essentially PR vaporware prior to these firings. Granted, there's a lot of engineering going on behind the scenes, but the timing is perhaps suggestive...

Look at the timeline:

Jan 2015 -- announce plans

Jan 2017 -- Starlink trademarked

Feb 2018 - test flight of two satellites

May 2019 - Launch 60 of the v0.9 model

June 2019 - fire 7 people

Oct 2019 - this article

Nov 2019 - Launch 60 of the v1.0 model

7 Jan 2020 - 60 satellites launched

29 Jan 2020 - 60 satellites launched

17 Feb 2020 - 60 satellites launched

18 Mar 2020 - 60 satellites launched

22 Apr 2020 - 60 satellites launched

04 Jun 2020 - 60 satellites launched

etc.

https://en.wikipedia.org/wiki/Starlink

I know that SpaceX has historically had a abnormally high development and deployment cadence, but I think that your second point about behind-the-scenes development is probably closest to the mark.

To me, it seems more plausible that they were taking a solid-not-polished design that's been mostly worked through for a few years, doing some team re-org and removing people who aren't working out, and then finished tweaking the design and got ready for scaling to higher production over the next few months than it does they were scraping by on the skin of their teeth for almost four years and suddenly managed to catch up on everything (and have it work!) in those same last few months.

It's absolutely possible I'm wrong, but if the Starklink team truly managed a timeline of 6 months from "vaporware" to functionally the world's first mega-constellation, I'd be absolutely floored.

If you just graph the launches that looks like R&D + production ramp-up.

Where do:

"Amazon representatives have had "30 meetings to oppose SpaceX" but "no meetings to authorize its own system""

... those 30 meetings fit into the timeline? :)

> Oct 2019 - this article

The article was published on Oct 31, 2018 (not 2019).