Let's not repeat history so soon

The way to get costs down for launching things to space is to make all the parts reusable, right? It's obvious, isn't it? If you can reuse things, the launches must become cheaper, surely!

How can someone even THINK that that might be true? If this were 1972, it might be excusable, but we have 30 years of history that say it isn't.

Several years ago, I attended a lecture at MIT by Aaron Cohen, the first Space Shuttle project manager. He finished the lecture by saying [I paraphrase], "If you notice, the Space Shuttle did not achieve a single one of its top level requirements. If I had it to do over again, I'd look at those requirements a lot harder."

That is wisdom in its hardest won, most valuable form. The way to design a system is not by saying, "Here's what we want it to be (e.g. reusable)." The way to design a system is to first state clearly, "Here's what we want it to DO." Someone's favorite approach may be incompatible with their objectives.

Another example from history: expendable launch vehicles were long designed to maximize the fraction of their launch weight that they could get into orbit. That makes sense, right? Well, maybe not, if you look at how much the fancy engineering to achieve that winds up costing--$5,000 a pound to low Earth orbit in the best cases.

Within the last decade, a few people, notably Elon Musk of SpaceX and some others, started asking a different question: "How do I design a launch vehicle for lowest cost per pound to orbit?" It turns out that the design is much different. Using this new paradigm, the changes in launch vehicle design have only started to appear on the scene. SpaceX's Falcon Heavy, when it achieves maturity, may give launch costs as low as $1,000 per pound to low Earth orbit (albeit only for huge payloads). Other approaches, not as far along, have visions of even lower costs, and for smaller payloads.

SpaceX has identified some savings that may be realized by recovery and reuse of their booster stages. I would never bet against them. But again, the vehicles were FIRST designed with low COST as the objective. Subsequent improvements to the overall business process are one thing; constraining the design to a certain approach with uncertain implications is another.