I don't follow. If you test 10% of a population for a disease where as many as 80% are asymptomatic, you'll get a certain number of cases, some serious and most not. Ifyou test 100% of the population you'll get a much larger number of cases but that doesn't mean that it's more than you'd have if you tested 100% of the population two months ago when it was peaking.
Hospitalizations is a better analog in my view. Those are serious cases which have a test result. Serious cases requiring hospitalization in March or April would have required it in June if this is when they occurred. So seeing how that number rises / falls is a better indicator than raw cases as I see it. Remember, the point of this was to prevent hospitals from getting overloaded. We're way past worrying about that, it wasn't an issue even at the peak. And yet we still have lockdowns.
% of positives of total tested gives more of a live action indication of how widespread Covid is in an area regardless of expanded testing. It is good to help in planning and containment that will hopefully reduce future hoslitalizatiins and deaths. Hospitalizations and death numbers certainly have their importance too but both lag considerably and relying on them alone let’s the virus get too far of a head start.