@Ego - No, you misunderstood the calculation in assumption 2. The 0.59% multiplier accounts for the 30-39 year old (it's the percentage of 30-39 dead out of all the dead). It goes in along with the 1.17% (the number of dead out of the total infected number). Please go through the math again if you want to argue this. This is likely just a slip. I doublechecked the probability factors when I wrote the post.
As for assumption 1, the 70% comes from R0=3.5 (herd im.=(R0-1)/R0) which is what the early estimates of the basic reproduction number were before people started modifying their behavior (resulting in a lower R0). The 3.5 number would account for kids and young people being immune and non-infectious and also account for the immunity of those with previous exposure to common coronaviruses. The reason is that the 3.5 is
experimentally measured based on early spreading rather than computed from detailed knowledge about children or antibodies. A SIR model example is given upthread for how it can be derived---all that's required is the growth in the death rate and how long it takes to die or recover.
R0 is initially independent of shadow numbers(*). The dynamics of the virus does not change because we now those these details. (Similar to how gravitational acceleration didn't change when Newton wrote his second law and measured g=9.8m/s^2.) It changed because we modified our behavior and because the fraction of recovered people increased.
(*) This is because there's only a 2-12 day latency between exposure and when people present at the hospital. It's not like HIV where the disease had 5-15 years to spread around before resulting in what was initially seen as mysterious symptoms. IOW, it's impossible to infect 10% or even 0.1% of the population with SARS-CoV2 before people start showing up at the hospitals.
(The 3.5 is just the average of early calculations. It ranged from 2.5 up to 7+ but most had it in the 3ish range. I'd be perfectly happy to use other reasonable numbers. I used 1.1 for Sweden because that seems to be their path goes forward. FWIW, the entire world is converging on R0=1. I explained why I think that is in an earlier post as well.)
The only contention here that changes the numbers is you second bullet point, that is, whether people fight off the virus w/o producing antibodies. (In other words, whether the Spanish study says something useful about the total infection rate.) If that's the case, then even more people have been infected. This then means that the IFR is lower and that we're further ahead towards the R0 based herd immunity; that it's not a factor 13 to go and that the eventual death count from going there will be lower. I do note that a factor 13 is a lot (way outside the ballpark of current experiences). It means that up to most people would fight off the virus w/o producing antibodies. I don't know how normal that is?
However, that's basically the weakest assumption in all these calculations. That there's a large shadow number out there. So the argument now is that antibodies don't measure [the shadow number] either and that some people are just naturally immune. Okay ... but if that's the position, there's really nothing I can say that can make a convincing point otherwise. There's any number of assumptions one can make that if true would mean the pandemic is largely over. However, I will note that the numbers that are currently coming out (3-4 months in) are very much in line with very early estimates. That these have not been falsified yet speaks well for the original models.
Anyway, I have no personal stake in the outcome because---cynical as I am---I've deliberate set myself up to be independent of the outcome. I'm prepared for things to go either way. I've long accepted that people are reluctant to accept anything that hasn't affected them personally yet (case in point, THF generally seems to want to learn things the hard way usually rejecting good advice when it doesn't fit what he wanted to hear---an attitude which I consider perfectly normal, maybe especially for younger people) and are not going to listen to the experts (
except when they eventually need help to deal with the consequences). Humanity's waffling reaction to CV19 has just confirmed that we're not up to the task of dealing with complex problems that require coordination at all levels of society. I consider
that my basic assumption. We're in over our heads. The ingenuity gap is here. As I mentioned earlier, if people can't even deal with the fallout of a pandemic economically or culturally---a kind of training wheel exercise for more serious 21st century challenges---there's no way they would be able to meet the 2C CO2 reduction goal before 2030. People are already ignoring that despite 40+ years of heads-up.
The majority of people can't even figure this out in their personal lives. The majority is steadily sailing into lifestyle diseases that will kill them but they're fine with that because they believe they're the exception. Experts can provide a pill for everything, but they can't fix stupid, so I'm pretty fatalistic as far as all this is considered. The main benefit of expertise is knowing what's coming. Not in preventing anything bad from happening. Sure, there will be a bigger pandemic at some point or even an actual food crises (as opposed to a temporary TP shortage). People will be warned ahead and proceed to ignore the warnings. This is all baked into the cake.
