It has been well documented that it is more efficient to build large power generation plants close to the source of the fuel and transmit the resulting power than to ship the raw fuel and generate locally with lower efficiency generators. Not to mention the significant difficulty involved in transporting large rivers for use in local generation schemes.
Some of the most promising work on future power generation involves tidal energy and geothermal recovery both of which demand a delivery grid.

The reason that the self contained wind/solar homes desire grid connection is that while the average power produced may meet or even exceed the house load the "time of generation" is not well matched to the "time of use" so a storage scheme is needed. Unfortunatly the grid usage curve generally follows the home use curve meaning the grid has excess cheap generating capacity available when the individual wants to dump power on to the grid and is ramping up expensive generators when the individual wants to draw it back.

Nuclear power... I wonder if anone has ever calculated the total energy used to construct,maintain,remove at end of life, store and render safe the radioactive waste created by and for the generation of fuel for the plant. Add them all up and compare the number to the amount of power generated.
Remember that some of the waste has a halve life longer than civilization has existed.
Do you think the answer will be more energy used to create the plant or more generated by it during it's 30 -50 year life cycle?
Terry