"Because there is no proof, let's use the lack of proof as proof."
Statistically significant correlation is actually considered valid supporting evidence, although not final proof. It helps shift the burden of finding a valid third variable to the party denying the causation. Of course, finding statistically significant results to support a hypothesized finding is very different from picking one significant correlation out of a data table.
With enough instances of a correlation, it’s likely that many would consider it proof pending a clear and obvious reason for dismissal.
What is a demonstrable third variable that applies to all locations that adhere to to masks/SD vs. similar and close locations that don’t behave this way that would explain the discrepancy in case load, and what makes you believe that variable is as or more likely than masks/SD?
You can find some interesting points about this
here