No story there. But then they draw almost flat lines, indicating the selectivity of the university (as a proxy for all kinds of things, prestige, quality of education/network etc) plays no role. But I find it misleading.
It makes sense I think to consider that they're not looking at earning potential, but at ROI. That means a selective school that costs $100k a year, and nets you a $100k annual salary, yields the same ROI as going to a 10k a year state school that gets you a 10k salary. You'd get a flat line on selectivity/ROI, as if it doesn't matter where you study.
You can be the judge of whether you think those are equal or not, and whether you'd be indifferent because they have the same ROI, particularly when you can afford the investment or get the investment subsidised.