At a high level, it's a really cool build you've put together. You might consider writing down some of your methodology and assumptions in the blog post, as it's unclear from the post alone where those multi-billion figures come from. And someone could reasonably argue that there's overlap between the categories you've chosen; for instance, is a big part of the goodwill value of a NPM download realized in the improvements to recruitment and code quality? Even so, even if many of the assumptions are inaccurate, the order of magnitude may very well be correct... and if so, it does speak to the open-sourcing of React being a very valuable move for FB, even if payoff is spread out over time.

Taking a step back, I think you'll find a big segment of Hacker News really dislikes clickbait in the way you titled and structured the first blog post. It's a shame, because I think your devil's-advocate analysis deserves more attention. Certainly the followup deserves discussion.