As many public health experts and virologists are now pointing out (and as I am sure many at CDC have long realized), there's no way to know if there's community spread without actually testing for it. Current testing capacity is severely limited - which is baffling really, given that Korea is running like 6000 tests/ days now. Part of it is a logistical / technical screw-up:

https://www.politico.com/news/2020/02/20/cdc-coronavirus-116...

For example, if you're a doctor with a case of atypical pneumonia, there's no way to actually test a patient for it unless there's clear China (possibly Hubei!) travel history. And that could take 3-7 days. However, consider the Canadian ER doctor who a few days ago decided to ask for a test for a woman who had traveled back from Iran recently - turned out she had it, and it revealed to us unexpected, but very real vectors of transmission.

These technical issues have to be solved, otherwise there's no way to distinguish this from regular seasonal flu cases and pneumonias, until - like in Wuhan - the atypical cases bubble up. At that point, it's really too late to do much containment.