Additional of establishment probability in re-emergence

In the `demographic_turnover.Rmd` vignette, we currently plot $R=R_{eff}$, which gives expected dynamics (increase/decrease). However, if R is small, a single introduction is unlikely to cause a large outbreak (full methodology is given in {superspreading}). For a simple exponentially distributed infectious period (as in SIR model), the probability of a large epidemic is $1-1/R$. Hence we could add this as an extension of Figure 1 to highlight the additional shift in risk that happens as $R$ increases further above 1.

Working branch for this is `turnover-outbreak-risk`.