Across 261 matched pairs from the AI-READI cohort, lean adults and adults with obesity shared nearly identical mean glucose (122.3 versus 122.8 mg/dL) and HbA1c (5.73% versus 5.78%). Continuous glucose monitoring still separated them: obesity was associated with nighttime glucose (00:00-06:00) about 6.2 mg/dL higher and afternoon glucose (12:00-18:00) about 5.1 mg/dL lower. The pattern held when aligned to wearable-detected sleep onset, in healthy and prediabetic subgroups, and with mean-glucose matching tightened to 2 mg/dL. The same average can hide a different 24-hour shape. Because HbA1c and mean glucose blur timing, an adult with obesity could look metabolically identical to a lean peer on a standard blood test while carrying more overnight glucose exposure, which might reflect impaired nocturnal glucose handling or altered hepatic glucose output. Effect sizes are modest, roughly 3-6 mg/dL, and the clinical meaning is unproven; nothing here shows that this redistribution predicts diabetes, cardiovascular events or mortality. The cross-sectional design rules out causal claims, and meals, activity, sleep quality and medication timing were not controlled, so the afternoon-versus-night difference may partly reflect eating patterns. This is a preprint that has not been peer reviewed, and the findings may shift on review. It is incremental work that adds a timing dimension to CGM research; for now it argues for interpreting glucose data beyond averages, not for changing individual behavior.