That red line (generally declining since 1992) is heading toward negative territory for the first time since 1979. We can see the late 90's dot com bubble (burst in 2000) in the graph... but the last time the two trends were so far apart was 1984. Why are they so far apart now? That's what I'm seeing. Why do I suspect that current GDP numbers are inflated and don't reflect real "health" in the economy? Anybody with me on that?
To me, the red and blue curves part company around 1995.

1995 is the year the Internet went mainstream.
The first Netscape browser was released.
Microsoft released their Windows 95 operating system.
These two events revolutionized the way business did business.
Documents no loger had to be Fed Ex'ed overnight. They could be emailed. People could start chatting online. Less need to drive and fly to every meeting.

Maybe the Internet is the bigger reason why big red is trending down?

Interesting point. My impression is telecommuting isn't really much of a factor - not many people, not many days, usually give it up before long. I wouldn't expect the changes in freight delivery to have much impact since trucks are a small fraction of total miles. The lines definitely look a little different in the last decade. However, one can also argue the era after 1991 or so is a long period with no oil shocks, and that might have to do with it. Not saying you're wrong, but hard to tell.

It seems useful to investigate aviation - it's transactions and economic activity, but removes the land use factors that seem like they'd come into play a lot in the highway mileage.

We switched from direct flight to hub and spoke systems and milage went up for air flights. We could switch back if jet fuel prices went up again. More aircraft, more crew, more routes, less time in the air, at higher ticket prices.
Good point.
Transportation Energy Data Book says that aviation passenger miles grew at 4.8% annually from 1970- 2002 (ie significantly faster than GDP), but it was only 2.9% from 1992-2002 (pretty similar to GDP). I guess the earlier part of the curve was very much still in the adoption phase for airline traffic, and now maybe it has settled down to be similar to GDP.
There is really a trend break in the '90s onward. This is clearly seen the second graph (y-to-y changes). The volatility of vechicle miles is down and the growth rate is consistently smaller than before. On reason could be that the GDP data is flawed. The way it is calculated has really changed, so this may very well be true.

The problem here is that we don't really know what this driving really is, what mechanism connects it to the GDP. The share of industry in the GDP has diminished and the share of the trucking also. This could explain the diminishing volatility.

This might be interesting if we were trying to find a new leading index for the GDP and use it for trading. But what it is we are now trying to find out relating to the Peak Oil? Will we try to see how effectively the oil prices affect the fuel consumption in the US? Then the answer is that they affect it a little bit but not much so far.

But this is not a good question. The Peak Oil is a physical phenomenon. So we might ask, how high the oil prices will go  when the supply goes down? Cutting the demand is easy in this situation - less oil, less consumption. But the price is only for allocating oil. It is not directly related to the physical supply. And the US is not alone in the world. Its share of the world oil use is only a quarter.

Would we like to know how diminishing oil supply affects the US GDP? We know already that it affects it. What we don't really know, is what happens when the oil supply diminishes for a long time - when world oil is in permanent decline.

Stuart is of course concerned with the oil efficiency of the US economy. But this is not the quite right question. It will not tell how the economy will behave. Highway driving is only one part of it. The total energy situation and efficiency is here the most important factor. The oil consumption statistics show that oil has been substituted for some extent. Changing lifestyle is much easier than getting more energy. Natural gas is now the real scare. Just move to electric cars if you could find enough power for that. Here we have the problem.  

Natural gas is now the real scare. Just move to electric cars if you could find enough power for that. Here we have the problem.
Actually, that's the solution.  If you burn oil or gas in a stationary combined-cycle powerplant, you can get 55-60% efficiency out of it.  Burning fuel in a car engine is 20-25% efficient at best; the average is supposed to be around 17%.  Using electric vehicles could easily double (and potentially triple) the effective miles per gallon, even if all power still comes from oil.

Then there is AE.  Wind is cheap if intermittent, and solar is relatively costly but coming down.  Even if wind can only displace 30% of demand, that would still turn a 50% cut in motor fuel demand into a 65% cut, and a 60% cut into a 72% cut.  Once propulsion is electric you have a lot more ways both to boost efficiency and to substitute.