Re: nCoV 2019-2020 Outbreak
They very rarely have the denominator right unless it's something obvious like ebola and it's in a well-monitored region. So... all of these numbers are going to miss some, many, or most of the denominator. The number you're looking for is the IFR, infection fatality rate. These are almost always based on educated guesses from what I've read, but not always. Logically, that makes sense because you never can account for every infection.
I'm also guessing you're more likely to miss the denominator low if it's a mild diseases like the common cold. The more severe, the more likely you're going to get it closer to correct as people die or seek treatment.
So you have a couple of options:
1. Assume the CFR is low, which seems the least logical, but it's not out of the question as resources become taxed globally instead of regionally.
2. Assume the CFR is high. Which I'd guess is fairly consistently off compared to the IFR. Even at 2%, it's a huge number compared to the estimates of seasonal flu at 0.01-0.1%, which is probably high because not all cases are reported AND it's moderated by the fact that so many high-risk people are immunized AND we have Tamiflu. So it's not even comparable anyways since we don't have a vaccine for this virus and only experimental treatments. Maybe in 2-3 years the CFR is less than the flu's, maybe significantly, after we start vaccinating our most at-risk populations. On the other hand, 2% is mainly based on areas with "worse" health care systems than the western world. So perhaps as this spreads west, it drops.
3. Assume the CFR is right. In this case, at 2% is a little scary.
What happens if you're wrong:
1. You're over-prepared, probably waste resources, and you might make people question your credibility next time.
2. You risk under-preparing. Probably the worst consequences as resources become scarce, people start dying, and you've still lost credibility.
3. You're probably a little of both, but at least you're closer to being fully prepared.