Earlier quoted context omitted.
Doesn't this page answer all those questions? https://www.cdc.gov/flu/about/burden/how-cdc-estimates.htm
No! I dug through that page and the cited papers, and it absolutely does not have any of the answers to those questions. If you find the answers on that page, I'd love to read them. For example, they say: > The numbers of influenza illnesses were estimated from hospitalizations based on how many illnesses there are for every hospitalization, which was measured previously (5). Ok... "estimated from hospitalizations"?…
They estimated the number of illnesses by dividing the number of hospitalizations by a series of probabilities (the probabilities constitute a funnel from someone getting an infection to someone being hospitalized - for example the probability you actually seek hospital attention, the probability your test accurately detects influenza, etc). The probabilities were uniformly sampled from a range of plausible values (derived from their survey data). They did this process 10k times to arrive at an distribution over the number of illnesses (this is a Monte Carlo simulation).