Include age structure

[adamkucharski]

Currently CFR and under ascertainment calculations are done using average dynamics (i.e. cases and deaths aggregated across age groups). However, it is often useful to estimate age specific CFR, e.g. Russell et al (2020).

This could be included by adding post processing on delay adjusted events, as in the above Diamond Princess analysis, or from exported traveller data.

There could also be scope to include comparisons with wider meta-analysis-based approaches, e.g. the mpox analysis by @lwhittles et al.