Nobody is claiming that. It is the third line down on the grapic I copied out of the article. Line losses (which in the the transmission line, substation, distribution and end user transformers) are shown. But they are damn near insignificant compared to the losses in a thermal power plant, especially a coal fueled power plant that may consume 1/3 of the generated power (past the generator) to clean up the exhaust to an accepted level.
Yep. The limitations of the Carnot cycle ensure that. A typical modern gas/coal fired steam plant runs about 45% efficiency. The new nukes run about 36%. Combined cycle plants are the most efficient of all and they run about 60%. So at best, 40% of the energy you use to make heat to drive a generator goes out into the environment. Line losses are large, but are also insignificant compared to plant losses. We could spend a fortune on a superconducting grid to lower those losses, but it would never pay off with today's technology.
Bottoming cycles can be installed in the condensers and generate additional power using Freon turbines and maybe eke out another 4-5% efficiency gain, but again the capital cost of such bottoming cycle equipment and the maintenance costs are huge and guaranteed to never pay off, which is why no one has ever installed such systems on commercial plants, except as demonstrators on small college campus power plants (University of Texas).
The guys writing that article had what we call in the nuke biz a Loss Of Physics Accident (LOPA)