I'll answer the question for my state of Michigan, which is one with a governor facing partisan backlash over her decisions:
> - Do the "extraordinary measures" have to be proven effective?
That's how it's been handled here. We went through a phase of virtually absolute lockdown this spring, loosened restrictions this summer, and locked a few things back down again (bars were especially bad). The governor also ramped up mask mandates as we learned more about their effectiveness.
> Do the "extraordinary measures" have to be proven effective than doing nothing?
Not sure exactly what you mean here, but the regions that have done nothing have seen worse results.
> Has anyone announced what metrics we're even tracking to determine "effective"?
Number of confirmed cases, number of tests administered, percentage of tests being positive, number of deaths. The last metric is the one we really care about the most, but we can mitigate it by tracking leading indicators. Nothing more than statistical theory at play here.
> How long does something have to persist until it's no longer an "extraordinary situation"?
> Are the "extraordinary measures" persistent or limited by time, risk, etc?
Michigan has a phased approach based on various metrics and indicators: https://www.michigan.gov/documents/whitmer/MI_SAFE_START_PLA...
I would love to see this sort of prudence for other crises.