Ran an unintentional experiment: three nights of 5-hour sleep during a deadline week, then sat my Diagnostic on Saturday.
Sustained attention held relatively steady. Reasoning dropped 2 points. Working memory dropped 9 points.
The institution's weighting of working memory at 20% of the composite means this single faculty drop pulled my Standing Δ down by nearly 2 points. The case report named it: "Working memory subscale regression of 9 points; consistent with acute sleep debt."
The off-screen curriculum is not optional. Sleep is load-bearing infrastructure for the faculties this institution measures.