I'd worry a bit about the summary of what is being reported.
> Facebook and Instagram deactivation improved emotional state index by 0.060 standard deviations (p < 0.001)
The link didn't click through to the appendix. This seems off, as small effects (the small number of standard deviations) tend to be associated with undesirable high p-values, not low ones. Though also, the 0.060 itself seems lower than the visual graphs indicate.