The COVID-19 pandemic

[tomh009]

That's not quite correct. She did not estimate it at 4000.

"What's more, Etches said computer modelling suggests as many as 4,000 people could be infected across the city."

That indicates a considerably lower confidence level. In the end, no one can say how many people are infected until we actually start doing some random testing. And that's unlikely to happen until things calm down.

That's exactly right: No one can say how many people are infected at the moment. But it is certainly higher than the confirmed numbers, and might not even bear much relationship to our confirmed numbers across time. And comparing our numbers to other jurisdictions' is pointless when no one is conducting mass testing, and testing protocols vary widely.

I agree that comparing absolute numbers has limited value, although it's clear that we have more infected people than Taiwan and fewer than Italy.

But I will argue that the growth rates are still relevant, as long as the reporting and testing methodologies stay reasonably constant. Because a 2.5% daily growth in cases, while still exponential, can be manageable, while 25% will likely stretch a society to beyond its capacity to deal with it. And 30-40% growth rates simply indicate that the situation is out of control. And the rate of positive tests (for example, 2% vs 20%) is also relevant, as long as the testing methodology is also considered.