> And the math there would no doubt be really interesting to an insurance company since you have the possibility that the launch is a success and the first stage lands, the launch is a success and the first stage crashes, the launch effectively fails (second stage failure) but the first stage successfully returns, and both stages are lost. That is a number of different outcomes to insure.
The insurance industry already deals with far more complex scenarios. A ship on an ocean voyage will often have the hull and cargo insured separately, different loss layers insured separately (e.g. first 10% of losses, 10-20%, 20-100% all separate), with multiple underwriters on each layer (and sometimes the same underwriter on multiple stamps).