Here is the best exposition of SLS I know:
Few people know about this article. I saw it on NSF over a year ago. However, I don't fully agree with some of its viewpoints either.
Starship's orbital refueling approach is not the future of interstellar travel. The required refueling volume is too massive, zero evaporation storage is impractical. Fuel is difficult to obtain at low cost from extraterrestrial sources (celestial bodies beyond Earth).
We should view future interstellar travel through the lens of modern transportation concepts.
The idea of using a single spacecraft to handle all missions is absurd. Interstellar exploration inherently requires multiple specialized transportation modes each optimized for specific tasks.
Now Starship has distorted this into forcing one vehicle to handle everything.
This resembles the F-35 situation - pursuing forced multi-service commonality resulted in a problematic product.
True large-scale deep space exploration should involve hybrid transportation systems combining different spacecraft types.
Different spacecraft should be reused across various orbital transfer routes in space.
For example, China doesn't need to obsess over deep space reuse of secondary spacecraft. Starship's second stage is optimized for escaping Earth's gravity well, requiring massive Earth-based infrastructure for its enormous refueling demands. Such oversized vehicles are unnecessary and unsuitable for lunar/Martian surface operations.
Real deep space vehicle reuse isn't implemented this way - it's extremely foolish. In fact, the US already has systematic research in this area. So does China.
An example: China's nuclear-powered dual-mode propulsion system (nuclear thermal propulsion + nuclear electric propulsion), currently achieving maximum thrust of 110 kN level (11 tons).
3-4 clustered dual-mode engines could reach Mars orbit from Earth orbit in 3-6 months. Single-mode (electric propulsion only) could reach lunar orbit in 6-10 months.
In Earth-Moon orbital transfers, fuel consumption remains minimal - just a few tons at most can deliver dozens of tons of pure cargo to lunar orbit (current chemical propulsion limit is about 1:1 mass ratio for payload vs transportation hardware). Starship's approach requires 14-24 tanker ships (essentially 2 refueling missions) moving thousands of tons of propellant (1,200-2,500t), yet only delivers 20 tons to lunar surface.
Electric thrusters (specific impulse about 5000s) require longer duration, but just a few tons of fuel can deliver 40-70t cargo to lunar orbit, enabling 15-20t of surface payload (excluding lander mass). Both orbital vehicles and landers in this example can be reused.
In reality, only the initial launch requires CZ9-class rockets to deliver 100+ ton spacecraft to orbit. Subsequent missions can use CZ10-class rockets for fuel/material resupply and hardware reuse.
Not like Starship's approach requiring multiple 100-ton fuel transfers and ultimately a 1000-ton fuel replenishment - all carrying risks and costs. Moreover, Starship's excessive dry mass leaves minimal payload capacity, with 90-98% of fuel consumed by the massive vehicle structure (waste increases exponentially for deep space round trips). The entire concept becomes meaningless.
