That seems beside the point OP was trying to make. They were suggesting random sampled tests in current hotspot areas where there's already a shortage in availa

That seems beside the point OP was trying to make. They were suggesting random sampled tests in current hotspot areas where there's already a shortage in available tests. I'm not sure what benefit random sampling would give you over testing actual potential disease candidates through sensible contact tracing in these areas. That might get you data slightly biased towards a sensible worst case and you're free to use data from other areas for capacity planning to compensate until they ramp up test availability.