clarify how {cfr} works with numeric example

from translation review by @Karim-Mane

how method works in spoiler callout as in slides

https://github.com/epiverse-trace/tutorials-middle/blob/a7cd4328ab3a594f79f98542ccf67452659d8817/episodes/severity-static.Rmd#L443

sample text

> in the early stage of an epidemic, the number of observed deaths is equal to the proportion of cases that will die (the true CFR) times the number of cases with an expected outcome by the day of analysis. (some cases will take some days to get a known outcome)
> When we account for this time from primary to secondary observation (from onset to death), an calculate CFR from the fraction of cases with expected outcomes, then we get a better unbiased or corrected or delay-adjusted CFR estimate.