This is false, energy efficiencies do decrease energy use, in most situations. Jevon's paradox is the exception, not the rule.
For example, increasing fuel efficiency of cars does decrease energy use, because fuel cost is not the limiting factor foe most car travel. Similarly, more efficient lighting, more efficient home space heating and water heating, and better weatherization of homes, all increase energy efficiency and decrease energy use.
There's only so much lighting I would ever want to use, only so much heat, only so much time I want to spend in a car. Energy efficiency is one of the most powerful tools we have to reduce pollution, because it makes all the other things easier.
As an anecdote: if you take an ant population (or any controllable population), and put it into a confined space with constrained resources, the population eventually dies due to one of two things:
1. Elimination of resources
2. Excessive waste product (in this case, excrement and bodily wastes; pollution)
You seem to be talking about the need of a species to rely on a greater ecosystem in order to survive. However, I'm not sure how that relates to demand elasticity and energy efficiency. Perhaps you could be a bit more explicit?
Yeah, and if the sun turns off and stops pumping 174 PW into our system, we'd be pretty screwed. Luckily we do not live in a confined space with constrained resources.
There's a lot going on with US gasoline consumption:
1) land use decisions are pushing people further from workplaces
2) consumers are switching from the "car" class of vehicle to "light truck" which is negating fuel efficiency improvements [1]. There's a lot going on here, from people getting larger vehicles for perceived safety, to better comfort, to just having more money to spend on larger vehicles. Plus, all those people who think they need a truck despite using the bed once or twice per year.
3) Demand for gasoline is incredibly inelastic in response to price, with huge price spikes having almost no change in gallons of gasoline even as consumption in dollars spike [2]
MPG is a measure of efficiency, so it makes sense that it increases. To support your argument, you would need to demonstrate a corresponding increase in vehicle miles traveled per car, so that total gallons of fuel burned (i.e. energy) is constant or increasing.
I agree more with the second point; the way that I'd say that is humans as a species will never self-moderate energy use.