Well, please note that the entire of U.S.A has tested like 500~ people in total. Some expect there to be many more undetected cases that we're missing simply because we are not testing.
That’s due to bayes theorem. If you test people who aren’t like to have caught the disease, and say the test has 1% chance to misdiagnose, you’ll almost certainly create wild amounts of false positives and spark a wider panic.
Sure if you want to cherry pick explanations. Likelier alternative is this is a new virus and the test needs to be adapted for it. Also we don't know where to swab for it. Another paper came out saying presymptomatic patients have concentrations of virus in their sinuses IIRC.