If we want more nuclear capacity we need to be serious about it. I think LFTR probably deserves a few billions in research, if it pans out it could make nuclear competitive on price and reduce a whole crap-ton of issues normally associated with nuclear.
Solar, though, is not a panacea, for a variety of reasons: http://www.reddit.com/r/politics/comments/243plj/the_koch_at... or http://www.newyorker.com/online/blogs/elements/2013/09/jcap-... or http://www.forbes.com/sites/christopherhelman/2014/04/24/sol...
Rather than firing up the peak-demand plants when load rises, just keep the plant humming along at the same rate and throttle back the cement producer, or biomass gasifier, or fertilizer plant, or aluminum smelter, or paper mill, or desalinator, until the grid load comes back down. Any industry that can singlehandedly soak up all the excess power on additional production would do.
Also, most if not all industrial electricity users have load shedding agreements with utilities; in the event of demand spikes, those industrial users are required to curb their power use within X number of minutes.
Energy storage, using electric car batteries as grid smoothing utilities is an interesting approach. Grid sized flow batteries[1] are another interesting option.
* Increased incentives for solar (which serves as a consistent energy source during daylight hours)
* Guaranteed rates for base load nuclear generation
* A carbon tax on fossil fuel generation, used to subsidize nuclear
Across the US, we already pay for reserve capacity, known as "peaker plants". They are paid simply to sit idle all year for those few days when we reach maximum system demand. It would be similar to paying low/zero-carbon base load plants (nuclear) to produce flat out, acknowledging their fixed costs.
I cringe at the thought of a substantial number of natural gas turbine plants being built, and the price of natural gas skyrocketing.