A lunar propellant economy is not one machine. It is a dependency graph whose edges carry power, mass, schedule, and institutional responsibility. Prospecting has to reduce uncertainty before a site hardens. Power has to survive before excavation can become routine. Water has to become matched cryogens. A reusable tanker and depot have to preserve that mass through storage, flight, and transfer. One tonne is deliberately small: large enough to force the complete chain to close, small enough to keep the first campaign inspectable.
The cost is conditional on the accounting boundary shown in the instrument. Buying a commercial service and funding the enabling stack are different questions, so they are never collapsed into one apparently precise answer. Expected mode is a deterministic baseline. Seeded risk is one reproducible history. The ensemble is an uncertainty experiment over the same event graph. None is a programme forecast.
EML1 is computed from a circular restricted three-body equilibrium, but the transfer arcs are illustrative. The campaign omits crew systems, full trajectory propagation, mature commodity markets, a second depot, and manual infrastructure placement. The evidence tab records what is observed, analogue-derived, or stipulated.