I want to provide a different view on tax policy here, which is that it is not logical and consistent, nor should it be.
I know the idea it should be is a popular perception among tax lawyers and economists, but our system of government does not put tax lawyers and economists in charge, and that's on purpose.
The reality is that tax policy exists to create specific outcomes in society--either by funding public expenditures, or by creating incentives to shape private expenditures. Politics is how we set and prioritize social outcomes. Tax policy comes out of a political process.
So as you evaluate a change in tax policy, you need to start with what you like and want in society. There's a lot to recommend graduate education, and in particular STEM graduate education (which tends to get the most tuition reimbursement). You don't detect gravity waves, find a vaccine for cervical cancer, or increase solar panel efficiencies without a lot of highly educated people. You don't attract the world's greatest minds without a prestigious, well-funded, and diverse higher education system.
Outcome-oriented thinking helps avoid unnecessary blinders about the scope of the tax system. We don't have to choose between STEM grad students and the EITC. There's a whole huge world of other tax policies out there we can think about.