I think this misses a key phase in technological progress which <i>doesn't</i> match simulated annealing:<p>Once you get to that final minimum, where you can only iterate to improve, people see it as the every-day, and start latching onto a new round of fluid creation (and there's the fight-The-Man attitude, too).<p>I'll call the growth-after-optimum a "flowering" period, where things shoot off in all directions. Frequently, they're based on that optimum, but not always (seeds spread).<p>(and how exactly is the Windows / Command key a major innovation? there have been function keys longer than there have been computer keyboards)
I can't find the source right now, but I remember reading a paper where it was assessed that a new technology needs to be roughly twice as good as what it's replacing if it is to be successful due to the inertia the post describes.<p>The post also makes you realise that marketing and evangelizing is a very important aspect of doing a startup, especially if your product is groundbreaking and threatening to the status qou.