I would argue that the transition from feudal societies to modern nations is a example for such an transition.
On a much shorter timescale, there seems to be quite some variability in corruption[1]. And connecting the plot lines, there seems no show stopper at any value of this corruption metric, so there seems to be no problem in principle to move from a very corrupt society to a uncorrupt one on a timescale of 20 years.
Playing a bit further with the data, the correlation between corruption and GDP is rather impressive [2].
Agreed, I was thinking that as well. I wonder if the same factors in transforming those societies all those years ago could be used in the modern world? Or has the world changed so much since then that the same factors could no longer apply?
Things like much larger populations, greater mobility and communications may make the solution to corruption a lot different today than what worked in the past.
Playing a bit further with the data, the correlation between corruption and GDP is rather impressive [2].
[1] Google public data explorer link: http://bit.ly/PQ12bj
[2] Again Google: http://bit.ly/Ttp24m