A lot of areas in the country are on the steep part (or the steepening part, hopefully it doesn't get any steeper than this) of the gompertz curve.
(Gompertz is an S-shaped curve that was too complicated to explain back in the spring and so the media went with "exponential" instead, a word which is more familiar to most people and sounds more impressive anyway... never mind that another key concept, that "flatten the curve" doesn't mean fewer total infections, was lost on 90% of the population- many of whom were probably busy eating paste during high school math and science classes instead of paying attention.)
I do agree with the general theory that
@Treetop Flyer mentioned, that there are a lot of cases out there of people who were infected and suffered only mildly or didn't know it at all. As to what that number is compared to the known cases, I think that's anybody's guess because I haven't seen any studies on random sampling of the population. COVID tests for anyone who wants one are easy to get in some places and still a big hassle in others (the only surefire way I know is to donate blood). I suspect the number of people who had it and got better is significant although I'm not willing to bet it's as high as 4~5 to 1. It'll be interesting to revisit this point a few years from now.