1) and 2) are the unavoidable consequences of exponential (actually, geometric) growth. But be reassured, this type of rapid, accelerating increase never continues indefinitely in natural systems. Sooner or later, it collapses. It never just stabilizes and levels off.

3) is a bit hard to believe. The Carnot efficiency of an internal combustion engine is the same as any heat engine operating in that temperature range, about 20%. Normal departures from optimum theoretical efficiency such as friction and heat loss probably account for about half again that. So my guess is the energy efficiency of the standard auto is probably around 10%. And remember, there are unavoidable losses too, a car rolling down the highway is consuming fuel solely to overcome friction. Theoretically, only accelleration actually requires energy.

However, if you factor in all the other indirect factors which contribute to automotive inefficiency (like being more powerful than they need to be, idling losses in traffic, energy consumed in mining, building, transporting, refining and paving for them, infrastructure costs and medical costs to treat the casualties and repair damages from accidents, the fact we use cars when we don't need them and use them to justify living, working and shopping further than we need to, etc.) the 1% figure might be justifiable. But still, that's really abusing the statistics.

4) As for carbon fiber, I don't know anything about its use in cars, but in high-performance boats (where it is used where strength/weight ratios trump costs, it is extremely expensive and used only where money is no object. Of course, those guys saw the handles off toothbrushes to save weight, and their motto is "if it won't break it's too heavy".