Forecasted ROIs are meaningless considering major is more important for ROI than where you went to school. Harvey Mudd isn't a better university than Harvard, Yale, or Princeton just because its graduates go almost exclusively into what is currently a lucrative field. (MIT, Caltech, and Harvey Mudd are good because of their rigor, which is unrelated.) As it stands, a forecasted ROI ranking would almost certainly be topped by miscellaneous dev bootcamps because of their cost structure.
I've never been a fan of the U.S. News et al college ranking methodologies. Publishing historical–as well as forecasted–ROIs would be a far superior metric.