I think I figured Coronavirus. Consider this well known DE that describes exponential growth:
$\frac{dC}{dt}=rC$. But few people know a more general form of it: $\frac{dC}{dt}=kC^p$.
It's just if p=1 then the solution is the exponential function. But what happens if p<1? Well then the solution is a polynomial of degree m
such that $p = 1 - \frac{1}{m}$. For example if p=1/2 it is a quadratic polynomial. If p=2/3 it is a cubic polynomial, etc.
It is very hard (or maybe even impossible) to prove though but trust me, this is true.

Now what does it have to do with coronavirus? Well in many countries the cases started to grow exponentially with a certain accelerating
(or decelerating) factor r, but then the trend gradually degraded into a polynomial function $C(t) = (\frac{r}{m}t+A)^m$.
So I believe in my county (the US) we started exponentially but if you take a look at a recent graph of daily new cases, you will see that it
looks more like a linear function. What does that tell you? Well a straight line is the *derivative* of a quadratic function.
So the cumulative graph is most likely a quadratic polynomial at this point even though it may be hard to tell due to the scaling, etc.

Attaching the graph so you can see the difference. Obviously if you tweak the scaling factors for both functions you can make them look very similar *locally* depending on the time frame.