Umm... he is arguing that CS is exactly what people should be focusing because CS (ML/data processing, etc.) applied to existing fields and data could lead to new insights - so more smart people going to CS is a good thing (at least according to Yegge).
My point is that CS has practically nothing to do with writing JS frameworks and frankly I doubt that most people doing JS development are the kind of people that could make significant contributions to CS or mentioned sciences - maybe a fraction of the top % - but JS is a low barrier to entry, easy to get simple things done, sufficient to solve real world problems like automating CRUD TPS reports - because someone needs to do that as well. I've had the misfortune of working with that platform - JS is by far the lowest quality ecosystem I've seen* (I never really got into PHP so I can't compare).
IMO Yegge makes a stronger point suggesting that data science people are being driven to "catbook" projects instead of "important stuff" - but those people are not your everyday JS developers.