Agreed on the shifting the consumption down the line. However, the primary energy advantage of Uber-like car services is that it reduces the high cost of energy to produce a car by reducing the pool of cars through optimizing their use (meaning, instead of a majority of cars sitting around doing nothing, using a pool of cars working most of the time would mean fewer cars needed overall). If all cars end up being full time service vehicles (through self driving?), one car would hopefully be able to service 20 people and replace all those cars. If the carbon foot print of gas/electric consumption is 12% of a car's total (the rest going to manufacture), then moving to a full shared service model would be a huge gain (and loss of the auto industry)