The part about rate of improvement following an S-shape is interesting and I'd like to see it empirically researched, but the fact that the last exemplar of a given technology is always the most advanced is quasi tautological. Why would you build a new plane if it wasn't better (in some way) than what came before it? Or a new computer/phone?