That's also a problem with "citation needed" argument - people who are in position to publish papers are often motivated politically or just bad at thinking and logic.
However I think that simply dumping money into education as it is is more likely to produce bad results than good ones. That's because the current education system is driven more by ideology than pseudoscience than by anything resembling an effective methodology. Given the "citation needed" atmosphere, I'll offer https://calteches.library.caltech.edu/51/2/CargoCult.htm as an argument. We have lots of changes being made educationally which are promised to produce better results over time. And, time and again, they don't. As Feynman said, the planes aren't landing.
But if we can invest, I also question the promised savings in the article. It does not distinguish between increase in income due to having the characteristics that help one get a degree, versus increase in income due to what is taught in a degree. This can be a very large difference. If you have a subscription to The Economist, read https://www.economist.com/graphic-detail/2015/10/29/our-firs... for how much college rankings and benefits change when you try to measure universities by estimates of value added.
That isn't to say that there is a better figure available for education. But it does not speak well to their intellectual integrity that they failed to point out this major shortcoming in their own data.
But the bigger issue is framing with as a “struggling vs gifted” problem. Where we have to choose between supporting gifted students or helping struggling students. It isn’t, education is one of the few areas where there is a “free lunch”. Every dollar invested in education results in more than one dollar in return. We can easily find both of these things, the real question isn’t which we should fund, but why we aren’t funding both.
[1] https://www.elc-pa.org/wp-content/uploads/2011/06/BestInvest...