Terraforming Mars proposals usually focus on the desired endpoint — an Earth-like atmosphere, liquid water, breathable air — and argue backward to what would be needed. The physics of warming is well-understood. The question that matters is industrial: can we move enough material fast enough?
Dong et al. work the problem from the constraint side. Achieving human-relevant atmospheric pressure requires inventories on the order of 10^{17}–10^{18} kilograms. Available CO₂ provides less than 20 millibars — minimal warming. Reaching habitable temperatures demands an effective infrared opacity target of 2–4, achievable through synthetic greenhouse gases or orbital mirrors but not CO₂ alone. And oxygenation — making the atmosphere breathable — requires over 10^{25} joules of work, demanding sustained industrial throughput of 10^{7}–10^{8} kilograms per second for centuries.
The useful finding is what doesn't scale. Regional paraterraforming — enclosed or locally warmed habitats — is plausible on near-term industrial scales. Planetary-scale atmospheric modification is not. The gap is not in physics or chemistry but in logistics: the throughput rate required to transform a planetary atmosphere exceeds any industrial capacity achievable without centuries of compound growth. Available volatile inventories are the binding constraint, not energy or technology.
The through-claim is about the unit of analysis. Terraforming discussions focus on the thermodynamics — can we warm Mars? — which is the solvable part. The unsolvable part is mass throughput over time: mining, processing, and deploying 10^{18} kilograms while atmospheric loss simultaneously removes what you've added. The bottleneck is not whether the physics works but whether the factory is big enough.