Global / robustness view
Instead of one set of inputs, this sweeps the group-level probability (conjunctive failure / disjunctive success / base case) across a grid and aggregates outcomes over the whole assumption space — showing how robust the outcome distribution is to your assumptions, not a single prediction. A uniform sweep assumes a uniform prior over each group's probability (itself an assumption). Normal weights the sweep by a Gaussian per group with a mean you set.