2) Surely this is redundant: most simulations already allow users to adjust variables and initial conditions themselves.

ANSWER:

a. Assumptions may be embedded in the fundamental design, such that the "dice are loaded" towards producing one particular result (eg Von Neumann's zero-sum assumption in game theory).

b. Even if different results are possible, the initial results reported will be the ones that count most in popular and professional communications.

c. Even if different results are possible, the results reported by professionals will be the ones that count most in popular and professional communications.

d. Even if different results are possible, stakeholders may not have access to the simulation itself.