That 27.8 loss in eletrical system losses is a freaking shame. We should be researching room temp superconductors as a national strategic project.
That's been going on since the early '80s. Problem is, high temperature superconductors are (as far as I know) all ceramics.
We already have it : a room temperature superconductor that converts ambient heat into DC current. It was announced at a materials conference, then immediately shut down on orders from the DOE.
Most of the loss in in generation. After power leaves the generation station system loss typically run 6~7% for transmission, substations and distribution losses combined.
Supply and Disposition of Electricity, 1995 through 2006
http://www.eia.doe.gov/cneaf/electricity/epa/epates2.html
What the numbers of 27.8 reflect is the amount of BTU's contain in coal (for example) that are delivered as power. Not only is their significant losses in total heat recovery, scrubber systems and utilities at a clean coal plant can consume large amounts of the total electrical power generated at the plant.
The Need for Additional U.S. Coal-Fired Power Plants
http://www.asme.org/NewsPublicPolicy/GovRelations/PositionStatements/Need_Additional_US_CoalFired.cfm
Most of the U.S. coal-fired plants are 30-40 years old and have been upgraded to meet emissions regulations. However, only about 25 percent of these older plants have been retrofitted with scrubbers, and only a few plants are equipped with the most effective flue gas cleanup systems technology now available for new power plants. While the average net energy conversion efficiency of these plants is about 32 percent, the next generation of new plants can increase this net efficiency to 40 percent or higher with improved technology, thereby reducing coal consumption and emissions in proportion.