>>13644120These particular type of model can't really be mistuned, unlike many others. They are data reliant and so they are only as accurate as the available data. At least, as far as I am aware, they haven't built a model that could extrapolate disease spread from the spread of other diseases. This remains true even with similar diseases to covid like SARS and MERS. R values are calculations on direct data related to the virus. Inaccurate early data on Covid, through whatever means, would yield inaccurate R values.
The other factor is building the societal consequence of some particular disease based on characteristics and the R value. Diseases spread through city centers different from rural areas even though transmission mechanisms, the disease itself, is not changed. And similarly, diseases have a variety of transmission vectors. Theoretically these could be used to determine R values, but there is a large black box in the immune system, immune variability, and virus behaviors which would have to be solved before it could be anything more than a guess. Population models are built on data points, infection rates, recovery rates, etc. The extrapolation between countries is finding comparable cities to approximate impact through profiling. I am not sure how they control for unreliable data.
Compared to mask models and 'droplets', which are essentially bullshit, the disease stuff is data driven.
