Let's review a few commonly-understood facts.
1) Public education continues to take more taxpayer money while providing little to no improvement in "outcomes" for kids, whatever that means [insert your own definition]
2) For those kids lucky enough to learn enough to be employable, businesses are now picking up the slack, with on-the-job mentoring and training. Business does a lot of training in the U.S.
3) Blaming one group or another isn't useful. It's not the parents, or the taxpayers, or the teacher's unions, or the administrators. What we have here is a systemic problem. The system is structured such that it sucks the life out of the teachers and continues to consume more and more resources. It's a bad system.
4) Making the system worse is the fact that it's full of consultants and external experts -- all of whom have some different idea of just what the hell the education system is supposed to accomplish. This article was a good example of that. Great ideas -- I love the idea of emphasizing relationships -- but markedly different from "Teach Johnny to read and write" or "Educate Sue enough so that she can get a job" or "Amit needs to be prepared for college". We all have different goals here.
So the problem is clear: there is no defined problem the education system is trying to solve. Therefore there is no strategy that can be "best", there is no expense that is either productive or non-productive, and there is no advice that is more or less useful than any other.
Until this issue of definitions is solved, no further progress can be made. In fact, no further progress is possible. Yes, you can come to the table with some great ideas, and you can make little snide political attacks against those who have different worldviews. Maybe even the other professors will pat you on the back. But you're not helping things.
So we're left with the problem Socrates gave us: the definition of terms. Without that, I am not optimistic that progress can be made. And guess what? Once we have a common definition of what the schools are supposed to be doing, we have a test. And so we are back to where we started. No matter what you do, you're going to have to measure it if you want to have some meaningful discussion about how it's done. That's an epistemological conclusion, not a political one.