This article was quite sloppy and inaccurate.
First, it makes the common mistake of comparing a graduate student stipends are low, while failing to assign any value at all to tuition. Many Ph.D.s are supported by grants, and their tuition is covered and is not taxed. That doesn't mean the tuition has no value, nor that money isn't exchanged to pay for that tuition. Real money is used to pay for that, and it provides a number of benefits to the student including access to the campus and all associated resources (libraries, gyms, labs, events, technology, workspaces, etc.)
People working at McDonalds don't have access to those resources from their job, so it's unfair to not include that when calculating their total compensation. When you do, they come out a lot better than low level drug dealers, as this article tries to argue.
It goes on to argue how graduate students are precariously employed. Quite the opposite, Ph.D. students are usually funded for 3-4 years under a grant. At least in the US, I don't know of many other jobs with guaranteed 4 year contracts, especially entry level. Most places are at-will and you can be fired on the spot. Ph.D. students are regarded as students, and therefore often aren't fired when they make mistakes (even big ones), because it's expected that their role is to learn.
That's another thing, is that this article focuses primarily on how the Ph.D. student is shortchanged by not being paid (in dollars) as much as a low-level employee at a corporation, but it doesn't spend any time on how such employees have different expectations. Ask a Ph.D. student if they want more money, and they say "Hell yes!" tell them they have to wear a uniform and arrive at 7:00am to open the lab, and they're fired if they're late twice, and every second of their day will be scheduled and tracked, they will be suddenly less enthusiastic at the prospect of making a couple extra bucks.