Compare variants and sensitivity if I had 20% more users

Compare variants and run a sensitivity analysis: how would the result change if we had 20% more users in each variant? Show current vs projected confidence intervals. Add a short recommendation on whether to ship now or wait for more data.