Thanks for the insights. Any links for simple folk like myself to see what the right way might look like?
Also, I didn't really answer your question directly before, but SpaceX-style iteration in this arena is unlikely. Unlike launch, which has high margins, commercial supersonics would be competing by trying to "build a better mousetrap" where "better" means "faster, but more expensive to develop, operate, and has a smaller addressable market". Iteration could build design team competence, but iterating the design itself would be extremely expensive and wouldn't have enough potential payoff to merit the necessary $Billions of pre-production investment.
My naivete as a software guy is that the advancements and cost drop of computational modelling, ML/AI, and 3D printing could enable a small team of smart people the opportunity to "think outside the box".