"Covid-19 cases compared by continent" - all very mysterious

From the BBC:



The vertical scale is not adjusted for relative populations - Asia is about 60% of the world's population and Africa about 20%. So the number of cases per million is tiny compared to Europe, North America or South America, which are about 6% of world population each. Why?

So not only is there a huge differences in the number of cases per million population, the number of cases peaked in April in Europe; in April and again in August in North America; in August in Africa and South America; and is still climbing in Asia. Why?
-------------------------------------------------------
You can sort of explain the two distinct peaks in the USA. You have to refer to worldometers.info; for (part of) the explanation. The graphs for 'urban' states like New York or New Jersey look similar to Europe - huge peak in April, now tailing off - and the graphs for 'rural' states like Nebraska or South Dakota peaked in August, having started from a very low base.

So the April peak was the 'urban' states and the August peak was the 'rural' states. The USA might as well be two separate countries (something that no presidential candidate seems to grasp). Bayard's theory was that population density is the main driver of how rapidly the disease spreads, but that doesn't apply to Asia at all. Why does this generalisation apply to the USA but not to Asia?
-------------------------------------------------------
The other thing to note is the disease seems to be becoming rapidly less virulent, so the countries or continents which somehow managed to stave off and delay the first wave did exactly the right thing. In most European countries, there are about ten deaths per day now, as opposed to nearly a thousand per day back in April. Again, why? And why are we still supposed to be panicking? From Worldometers.info: