They want you to assume that it's everywhere now.
I've been carefully tracking COVID-19 numbers for MD and St Marys County since this started. I thought this morning that it would be interesting to estimate how many cases might actually be CURRENT - since the total numbers keep climbing but most of those people are over it already.
So I created the attached plot based on an assumption that most COVID-19 cases last 10 days, and based it on the number of new cases reported each day. Seems like St Marys and Calvert are actually doing pretty well. But Maryland as a whole is NOT; a similar statewide graph is still trending upward.
Still, although it's trending down, the answer still makes me inclined to be rather cautious when out and about.
If we assume 10 actual infections for every one tested infection (based on several analyses of recent widespread test data from California), I think we could assume that there are about 300 infectious people in St Marys, and 250 of them don't know it. Given our local population ~114,000, that's about 3 people in every 1000, or 1 in every 330. (Some mostly-discounted estimates said 50-80x, so it could be much higher than that.)
So EVERYWHERE? No. Chances of encountering one of them? Not trivial. If there are 150 people in Walmart or Lowes at any given time, there is a reasonable chance (1:2) one of them is infected and infectious and just doesn't know it.
Edited: tweaking numbers based on additional research into infection duration.