That article points to a study that supports her claim, testing healthcare workers to demonstrate that they weren't infected, not merely asymptomatic.

No Googling necessary. It's right there in the article you linked to.

There was data to suggest that. The data that suggested that was linked in the article you posted:

https://www.cdc.gov/mmwr/volumes/70/wr/mm7013e3.htm?s_cid=mm...

People sometimes speak in hyperbole as a point of emphasis.

Behold, humanity.

If we can now expect the leading health experts in the world to be potentially speaking in hyperbole, then we can't expect the public to trust what they say to be accurate.

That's a problem.

If you take someone casually rounding up from 90% to 100% as a deep reason for concern, I have some bad news for you regarding weather forecasts.

Surely, you can't be serious. The setting isn't a paper but a news interview, where they want the general idea of a study, not the details. In a news interview, saying "people who smoke get cancer" is perfectly acceptable (and not likely to "get people killed"), even though not 100% of them do.