I think it's fair to say you missed my point. The college example was used in pointing out a comparable situation, in the analytical model you were citing. If reducing in-school access to the young reduces overall educational attainment throughout life (which is the premise of your article), then so do other crimps on schooling, including unaffordable tuition, poor school meal nutrition, lack of early learning options, lack of social services to single parents and at-risk youth, etc. IOW, you can't have it both ways, by being nobly committed to in-school teaching while neglecting the other components of what the article calls educational attainment.
And I wasn't inventing quotes. Conservatives at this very forum have dismissed the value of progressive programs with those sorts of remarks. The point is that you can plead one progressive measure only if you rationally accept that education is a continuum of needs spanning the first two decades, and more, of life. If you do that, then perhaps my point about the limited and speculative character of that JAMA study will be clearer.