Read it - you seem most in need.
At this writing, Ice_Ice's link above is giving 89,692 as the total cases number, yielding 2979 deaths, worldwide. That's 3.3% total, worldwide. That's at 42,606 recovered. So that's an interesting metric if you think about it -- at the disease "half life" (roughly half recovered, half still infected or dead) you get 3.3% fatality. The
apparent way to look at this is misleading, since many are new, many are cycling out, and so on.
I think that both S. Korea's and Italy's numbers will still change but I don't know anything, beyond the fact that these are small and new clusters compared with the "mature" Chinese case.
There is likely a tendency not to count infection
or death as coronavirus related unless one has tested, since its symptoms are similar to those of other diseases.
It's been a few days since some headline stories on our testing being minimal here. I'd love to see comparative protocols on when you test, for example, spanning Italy, ROK, and U.S.
Think about it: If you find ALL cases through testing (i.e., thorough testing,) and deaths are noted/reported at a constant rate, the apparent fatality rate compared with all infections will plummet.
So in the U.S. case, somebody might figure out that ramping up testing is GOOD to prevent deaths (and therefore the
reporting of deaths), and also GOOD to water down the apparent lethality of the disease (as the full incidence of cases becomes larger compared with the constant report of deaths).
But knowing the number of
cases might be the bigger ding to the economy, e.g., events, hospitality industry, etc... even though knowing the number of
deaths at the same rate, the apparent death
rate increases the less you know about the infection rate/number.
The true horror is that if reliable knowledge of
deaths is unavailable, the rate will also be unreliable. That's when we'll be in full Soviet-style statistical vacuum/denial.
Lest we go full political, right now we'll stick with "the numbers that we know are the numbers until we know otherwise."
Here's the "scoreboard" link again, thanks to Ice_Ice
Operations Dashboard for ArcGIS