se2015. I agree positive rates are not a precise statistics. But it is an early predictor. Overall excess mortality would be a better measure – but it is available maybe a month later after the deaths, or 1.5 months after the first infections. You can go back over last several months and compare excess mortality with positive rates 1-2 months before – and see if positive rates predict mortality.
There are other ways to get early predictions – google searches for “I am coughing”, wifi thermometers, but I am afraid best predictor would be population activity level, especially inside (schools, bars, shuls), and percentage of people wearing masks …