I think this can be misleading in the same way some charter school results are. The easiest way to improve a school's results isn't to improve the education provided, it's to get rid of the worst performing kids. Charter schools do this by various selection effects, and artificial barriers, like ending at noon on a Wednesday. So the only kids who go there have two parents, one who probably is stay at home and can pic…
"For starters, families have access to housing and health care through the military, and at least one parent has a job."
"Prudence Carter, a Brown University sociologist who studies educational inequality, said the Defense Department’s results showed what could happen when all students were given the resources of a typical middle-class child: housing, health care, food, quality teachers."
The article author argues that multiple factors contribute to this outcome, without identifying a dominant factor. Are you arguing that the selection effects you describe are the dominant factor?