And it's not just us, but epidemiologists seem concerned as well.
I've been following Marc Lipsitch, an epidemiologist from Harvard, on Twitter to try to cut through to the experts' opinions.
To piggyback off of your post with some more numbers:
1. COVID-19 is more infectious than the flu, but there's still a lot of uncertainty on how far it'll spread. Lipsitch estimates 20-60% of the adult population will be infected. The range is wide because of uncertainty around the R0. He specifies adults because at the time of his writing, there weren't many cases involving kids.
2. You can't just multiply the above estimate by 2-3% because that's the estimate for the Case Fatality Rate (CFR). The Infected Fatality Rate (IFR) is certainly lower because of all the mild infections that go unreported.
3. There's a great deal of uncertainty over the CFR. The 3.4% estimate comes from WHO, but experts are also looking at South Korea's 0.6% CFR as a more accurate measure. The reason is South Korea is testing the crap out of everyone with 140,000 people tested and 6k cases found. And the more thorough the testing, the more accurate the true CFR is (as opposed to only testing severe cases).
I'm dumb when it comes to this stuff, but it seems to me that 0.6% is very much a lower bound. They get this by dividing 40 deaths by 6000 current cases. But when a) cases grow exponentially, and b) deaths lag by say a week, you're going to underestimate the CFR by quite a bit due to exponential denominator growth while the numerator bakes.
4. There's worry that this will overwhelm the healthcare system. The flu infects 10-50m people each year in the US. Lipsitch's estimates puts COVID-19 infecting 40-125m in the US. Are we ready for the load?
My dumb opinion is that Korea's 0.6% might also be low also because cases haven't saturated the system like in Wuhan. And that maybe we can slow down COVID-19 by social distancing so that never happens.
All in all, COVID-19 might infect 2-4x more people than the flu and kill at a rate 6-30x higher.
