The comparison between skilled and more unskilled labor is intentional; the article was about training in science. Many posters seem to argue below that science

The comparison between skilled and more unskilled labor is intentional; the article was about training in science. Many posters seem to argue below that science training does not lead to better-paying jobs, and thus it is not important that poverty and social background are barriers to science training/doing science. A science degree in general requires more training than a business degree or a communications degree, and students from lower socioeconomic backgrounds are disproportionately ending up in "general studies," communications, business, etc., which often lead to jobs like restaurant manager (and don't lead to admin jobs in science). Maybe the difference between how I see it and what you're pointing out in your two points is that "failing out" or stepping out of pure science research into science industry/admin preserves or increases pay, while "failing out" or stepping out of business or communications leads mostly to unskilled jobs which pay poorly. The expected value of a job after a science degree is higher, because there are more good paths (even if some don't pay super-well).