Another way to look at this is just as an optimization problem. We found a good local optimum and it became too costly to search further. In the early days of

Another way to look at this is just as an optimization problem. We found a good local optimum and it became too costly to search further. In the early days of aerospace, the search space was mostly unknown and there were tons of companies investing R&D in a wide variety of designs. Huge amounts of human capital were invested, and as in all things, some designs worked, some didn't. Many companies went bankrupt. Eventually we find some good designs and the risk/reward of searching the space further just doesn't make sense. Only recently with improved tech and ML/AI + simulation has the cost to search been reduced enough that it makes economic sense to try again.