You raise a strong point.

The testing has to continue, but as long as visitors are coming from countries with a covid ratio of less than one in 10,000 (actually, I have no idea what the ratio needs to be, but it's not hard to calculate with a simulation) then a test each week would catch and contain.

On the note of modeling or simulation: the simulations are easy to assemble. They're just distributions of contacts, durations of contact, and upper limits on probabilities of contagion. Then it's a matter of running the simulation to calculate upper limits on just this sort of parameter, "max probability of leakage."

But yeah, the testing doesn't stop until the world has beaten it back with treatment, virus, or somesuch.