A smart grid would mitigate (but not completely solve) the problem of intermittency of direct solar solar energy and indirect solar (wind and hydro) energy --- most especially of wind. Thus the need for storage, say via battery banks, would be reduced (but not completely eliminated). High voltage DC can be transmitted about 10 times further than AC (thousands of kilometers for high voltage DC as opposed to hundreds for AC). Wind energy integrated over such a large area, comparable to that of the continental United States, fluctuates much less than at one wind turbine or at one wind farm.
The Australian Smart Grid Smart City trial showed that a major benefit is improved knowledge at the connection points to allow the efficient valuation of generation and demand. In simple terms, improved tariffs allows distributed decision making.
In a recent paper (presented in poster form to the ENA conference), I argued that efficient tariffs would allow customer sites to choose between using energy, not using energy, storing energy or generating energy. I need a snappy abbreviation.
This would include the capital decisions to support those choices based on the simple premise that people that impose on a market should bear the costs and those that assist a market should gain the benefit.
In the presence of efficient tariffs/charges/payments market entities will emerge to exploit the pricing regimes to the benefit of the customer without central planning.