It doesn't take 20 years to do that, it takes 20 years to do that and wade through the bureaucratic morass. The SR-71 went from initiation to deployment in under a decade, more than half a century ago. With the myriad of advancements in everything from engineering, computation, to business development/management practices, building new cutting edge planes is the sort of thing we should be getting better and quicker at.
Design iteration cycle-times should be decreasing due to CAD, experimental cycles-times reduced due to the proliferation of rapid-turn 5-axis CNC mills, experimental cycles reduced due to simulation, business processes streamlined due to advancements in JIT manufacturing and six-sigma/kaizen/etc, and so on and so forth. That this isn't occurring is a giant blinking red light that something is wrong, and that we are going to get our lunch eaten by someone who researches, designs, and manufactures with a modicum of competence. Ostensibly China.
Design iteration cycle-times should be decreasing due to CAD, experimental cycles-times reduced due to the proliferation of rapid-turn 5-axis CNC mills, experimental cycles reduced due to simulation, business processes streamlined due to advancements in JIT manufacturing and six-sigma/kaizen/etc, and so on and so forth. That this isn't occurring is a giant blinking red light that something is wrong, and that we are going to get our lunch eaten by someone who researches, designs, and manufactures with a modicum of competence. Ostensibly China.