> e.g. if a high-end CPU has bad transistors in a section of the cache, that section of the cache is disabled and the CPU can carry on its life as a Celeron

For me, that's the part that is really fascinating about the fab industry. We have a statistically-reliable method of fabbing, but we build in the necessary fallbacks in the chip design itself that if a chip partially fails quality control, we can scale it back and sell it as lower-end product. It's intriguingly analogous to the quality process in gemstones, except that chips aren't found---they're made via mechanisms that we can't perfectly control.

The way I've heard this explained, fuel is 3% of the mass of a modern car, 45% of the mass of a passenger jet, and 90% of the mass of a rocket. That means the structural engineering margins of safety for a rocket are much smaller, as there's very little mass fraction available to build redundancy into. There just isn't a lot of 'rocket' in a rocket. That, not complexity, is the root cause of why rockets aren't so reliable.

There's an even more basic physical reason why 'rockets are hard'. It's easy to explain in energy units of kT, the approximate energy of a gas molecule at room temperature. A carbon-carbon bond has an energy of 140 kT, while the gravitational potential energy of a carbon atom on Earth is 300 kT. If the Earth's gravitational potential were much closer to kT, room-temperature gas molecules would be able to diffuse out, we would have lost our atmosphere to space, and humans wouldn't exist. But that means that for a chemical system to escape Earth's gravity, needs a clever arrangement whereby most of the energy contained in its chemical bonds goes to a small fraction of its molecules. In a way, it's surprising not that chemical rockets are hard, but that we're lucky enough to be able to use them at all.

> The number of components in rockets and IC transistors are probably roughly comparable.

Unless you are counting the payload, I really doubt it.

ICs are only possible because people create some way to mass manufacture those features, and people don't even design the features by themselves.

If people created them at the traditional route of "manually design|manufacture this -> encapsulate into component -> use components to manually design|manufacture that -> encapsulate again...", it would be impossible to create chips as complex as we have now. But there still aren't better way to produce rockets.

All the problem with rockets is that they are created by joining a big number of things that don't really like to stay together (or, in a few cases, don't like to stay by themselves either). I doubt they have more bare complexity than a modern car.