add challenge or discussion on the implications of the percent of observations in EpiNow2

[avallecam]

Challenge:

• compare different percents of observations %
• how are they different in the number of infections estimated?

Discussion:

• what are the implications of this change?

Examples:

• how to estimate the observation % of a given surveillance system? Look further into https://github.com/epiforecasts/EpiNow2/discussions/636
• what factors, actions or changes affect the observation % during an outbreak?
• Is the test positivity rate a way to measure this change? Is this an example of sampling bias affecting Rt estimates? In Pitzer et al, 2021 they adjust by the percent of individuals testing positive.

Reference documentation: Accounting for incomplete observation from Gosnic et al. 2020