Your assertion that there would be "some clustering by random chance" is correct, however rocket failures are most definitely not described well by a Poisson di

Your assertion that there would be "some clustering by random chance" is correct, however rocket failures are most definitely not described well by a Poisson distribution. The simplest reasonable model would be to assume each launch campaign has a fixed probability of failure, and that the probability of any given failure event is independent of the others. In other words, a binomial distribution. And this is a commonly applied model, I've seen a number of people doing such calculations in the last few days. However, these assumptions do not accurately reflect reality. When a failure occurs launches are stopped, the problem is identified, and the problem is fixed. A failure event should cause reliability to increase for subsequent launches; failures are not conditionally independent.