There is another way of thinking. When one barely have a shortage of something, like spring water, which flows from mountains (and fluctuates heavily depending on season, monsoon, etc) one probable doesn't store much and let it flow.
For wind and solar, it seems, this is much better paradigm. Send it to the grid when its available, get it from the grid when it is still and cloudy.
- Supply side economic incentives (heavily variable pricing)
- Accessible, standardised real-time and forecast information about rates
- Client side automated usage control (with defaults)
This could be a good application for IoT. For instance, the only use-case I'd seriously consider for an internet enabled fridge, would be if it could "store" energy in the form of deep-chilling a thermal insulator to be charged at lower rates, when it's windy say.