If, and I mean its not certain, they are trying to change perspective on ME and CFS toward psychogenic causation by comparison to LP then they are playing with fire. First, LP might do better than them. Second, seeing biases in SMILE might be easier, and from there its a short hop to seeing them in the CBT/GET studies. Its similar biases in play.
I regard it as extremely likely that any perceived useful effect is due to bias. The bias in LP might turn out to be worse than the one in this form of CBT. In other words, it might be more successful as bias.
We are far past the point where every study needs realistic objective outcomes. School attendance is only part of it. Assessment of grades and other activities, total activity as measured by an actometer, etc., should be required as measures. What good is attendance if, and this is hypothetical, they go from A students to F students? We need to know about not just participation but quality of that participation, and schools do regularly assess students, its part of education.