In the face of scope creep, changing requirements, unforeseeable external factors, and purely human factors (like corruption, incompetence, neglect, internal rivalries, etc.), projects always faced huge risks. Projects nowadays are vastly bigger in scope (perhaps unnecessarily so), but the usual mechanisms to control the inefficiencies of bureaucracies are not easy to scale up.
How are you sure there is a better method than the educated guess? We have sophisticated methods of guessing and risk mitigation but I don’t believe there’s a fundamentally different approach.
After 2,000 years surely we should have found a more realistic means of judging projects and success than the upfront guesstimate …