Another major flaw in logic in the article IMHO.

In 1965, world oil production was 12 billion barrels. It may peak soon at 30 billion. Estimates project that in 2040, production will have slipped to 12 billion barrels--back to 1965 levels. To descend to that point would require a drop in consumption of 2.2% per year for 35 years. Can we do this? I think so. From 1973 to 1975, and again from 1979 to 1983, consumption fell by roughly this much per year. When prices fell, consumption rose again. For a glimpse of the future, note that when gasoline prices briefly spiked 30% due to Hurricane Katrina, US usage dropped 6% over two weeks. Saving 2.2% each year is well within reach.

The math is not this simple.  There were a lot less people in the world in 1965 than today.  We have already made large gains in efficiency.  To decrease oil usage at the same time as maintaining or increasing population is going to require a lot more savings than 2.2% per year because most of that usage is not distributed equally among the worlds population.  This is assuming a fixed rate decline on the backside of peak, which I question as well.  

Try running a world of 6.5-9 Billion people on 12 billion barrels per year, the same as many fewer Billion people in 1965 and see how successful you are.  Something has to give quickly, either population declines or we get super efficient very quickly.  

The problem with banking on efficiency is that you get most of your gains early in the process.  Doubling the fuel economy of a 50mpg car doesn't save you near as much gas as doubling a 25 mpg car.  Same increase in efficiency big difference in consumption gain.