Launch mass is the
wrong denominator.

Transportation cost does not end at touchdown.

SpaceX currently lists Martian surface cargo flights no earlier than 2028 at a rate of $100 million per metric ton. That listing makes the first-order constraint unusually visible: hardware that arrives on Mars must create more useful mass than it consumes. A delivered wall is expensive. A delivered factory that can make hundreds of walls is a different economic object.

This is not an argument that launch price will remain fixed. It is an argument that in-situ production matters at any launch price. Lower transportation cost increases the amount of equipment that can arrive; it does not remove the value of turning local material into roads, shielding, berms, foundations and utility works.

An illustrative campaign—not a prediction.

Consider an 8-tonne surface system, including its core spares. Each campaign imports 12 tonnes of binder, wear parts and other consumables, then incorporates 180 tonnes of qualified local material. Two campaigns place 384 tonnes of infrastructure from 32 tonnes imported.

The numbers above are adjustable hypotheses. They exclude power-system mass, upstream excavation equipment, replacement intervals and mission-specific logistics unless those are included in F or C. A real ledger must include every dependency.

The metric changes the machine.

1. Repairability becomes payload performance.

   Every additional campaign spreads factory mass across more output. Replaceable wet paths, sealed joints and recoverable software states are economic requirements.

2. Mobility becomes utilization.

   A fixed machine can strand capacity at one site. A rovalizer can move from a landing-zone berm to a cable trench, shielding job or maintenance repair.

3. Material efficiency can beat peak speed.

   A slower process with low imported-binder demand may outperform a faster process whose chemistry must be shipped from Earth.

4. Evidence becomes the product.

   Installed tonnes only count when geometry, process history and acceptance tests show that the work is usable.

What would falsify the thesis?

Swovee should not go to Mars if its complete support architecture cannot exceed the infrastructure mass multiplier of simpler alternatives: unprocessed regolith berming, direct compaction, prefabricated imported panels, or purpose-built single-function machines. The comparison must include power, excavation, spares, operators and failed output—not only the rover.