[Note: Updated per observations/corrections in early comments. CFR has been changed to IFR and statements that 88% figure applied to people not knowing they have COVID-19 (including post title) were corrected.]
What might be missed here at first glance (and is somewhat glossed over in the NPR report which is more focused on the low death rate) is that if we know (or can reasonably estimate) the IFR, and we know (or can reasonably estimate) how many people have died from the disease, then we can arrive at a good estimate of something we currently don't know: how many people have actually been infected since this all began?
[Just to illustrate, here's a simple example: if we know 7 out of every 1,000 infected people die, and we know 14 have died, then we know 2,000 people have been infected (14 ÷ 7 x 1,000 = 2,000).]
As of yesterday (May 28th), 103,330 people in the USA are known to have died from COVID-19.**
103,330 ÷ 7 x 1,000 = 14,761,428
So, we can reasonably estimate that 14.76 million people have, or have had, COVID-19 here in the USA.
However, as of yesterday there have only been 1,768,461 confirmed cases, or about 1.77 million. There are a number of reasons why there are so few confirmed cases compared to our estimated actual cases of 14.76 million. A large number of those that are not confirmed are people who are asymptomatic or who only have mild symptoms and therefore never considered getting tested. There is, sadly, also a large number that have been symptomatic, and perhaps very ill, who were nevertheless not sick enough to require hospitalization and therefore never sought the kinds of medical assistance that would have resulted in being tested.
What the numbers can tell us is that confirmed cases are a small percentage of actual infections. From the above numbers, we can now estimate that “infected but never tested” number to be about 13 million. Put another way, 88% of infected people aren't tested and many are unaware or unsure they actually have COVID-19.
This is what makes re-opening the country so scary. We currently (as of yesterday) have 1,166,406 active, confirmed cases in the USA. But we now know that this is only around 12% of the actual number of cases (because 88% don't get confirmed). Therefore, there are 9,720,000 people coming out of lockdown and many are unaware they have COVID-19.
Please keep that in mind as you re-engage with your friends, associates and fellow community members.
____________________
Note: I’ve lost my reference/link to the 0.7% IFR number I’ve used above. However, it fits well within the 0.5% to 1.0% range for the IFR that appears in the NPR article. As I recall, the 0.7% figure came out of studies in Europe that were notable for their large sample size (“blanket” testing within communities based on locale rather than, for example, self-reported symptoms). I believe one was based on data pertaining to communities in Spain. Note: I’ve added a link to a YouTube video about the Spain Study in a comment, below.
Also worth noting: There are no 100% reliable figures available in the middle of a pandemic. The actual numbers of infections and deaths are likely to never be known. There are no shortage of statements to the effect that the actuals are much higher than what is currently reported. In all honesty, it may be many years before an “accurate” IFR is known. That having been said, I see no reason to think 0.7% is going to end up being significantly off.