Does this study change your mind on that 20-30% figure? After all, the study measured a 35% infection rate for people living in close quarters with other infected individuals. Surely, the overall population would yield a much lower infection rate than this?