I would take the groupings with a huge grain of salt. While I agree that many schools are at high risk of closing, and that the work should be commended for taking a more systematic approach for predicting which schools will close, I suspect that the data or inputs used for this prediction is full of poor assumptions or junky info.