Add state level estimates for South Africa

[seabbs]

Requires:

• [x] Adding data extraction to covidregionaldata (see https://github.com/epiforecasts/covidregionaldata/issues/107).
• [x] Exploring available data i.e are test positive cases, hospital admissions or deaths (or potentially several of these) of interest and available.
• [x] Adding to R/list/dataset-lists.R as for other examples but with updated options based on data used (as default is using cases).
• [ ] Identifying better estimates for delays from onset -> report/death than used globally. If none found use global default.
• [ ] Checking estimates when produced from the batch run (each night but may depend on scheduling).
• [ ] Adding to website as a default page (see https://github.com/epiforecasts/covid/issues/116).
• [ ] Checking over an extended period of time to check estimates are being produced and results output.

Maybe of interest @pearsonca.

[seabbs]

Some examples of usage in the wild are:

EpiEstim based: https://lrossouw.github.io/covid-19/estimating_r.html Using the Imperial approach: https://lrossouw.github.io/covid-19/modelling_covid-19_in_south_africa_at_a_provincial_level.html

[seabbs]

Done by @sbfn?