Any time you convert power to one source, to another, there is loss. Right now, most home renewables convert DC into batteries, which are potential chemical energy (that also includes degradation-as they lose charge with time).

Many hate the idea of using the existing base load generation loads for charging off-peak, but unless it's really significant drop in usage during the off-peak, many plants nearly run 24-7 at a 100% because they have to keep the water boiling to make the superheated steam to run the turbines. They cannot really back off the fire so to speak, like you would in a car to slowdown, because power demand goes up and down every moment of the day. Even in the off-peak hours, demands can shift around quickly as second and third shift manufacturing can change the pull of electricity across the transmission lines. So real output of steam is pretty constant, unless the power company knows they can shut down a whole unit safely.

With that in mind, if cars are pulling that excesses energy that is potentially there, it would save us from having to burn real gas during the day for commuters. Most of the coal would stay the same, even if every home went to charging cars because most will be at 220V, and will be long and constant (with electricity, the longer and more stable the LF, the more efficient the generator can be at predicting and fueling). Your net increase would be less than 10 percent overall increase if suddenly every household converted (caught that in an Energy Central Magazine). So it would actually make the electrical system more efficient, and do away with less burning of fossil fuels, even if coal plants are generating (because they are all ready running anyways).

You won't be getting significant amounts of power from wind energy without better energy storage technology than what we have now (which is pumped hydro, which is a useful addition to any grid where it can be used) come along (i.e. if you want to stop global warming NOW it's nuclear or nothing). There also isn't really any unused power, unused capacity sure, but the power plants do get throttled down when demand drops (of course different power plants respond better to load changes, hydro is pretty quick as is natural gas making those good for peaking (though the more efficient combined cycle natural gas isn't as quick to respond) while coal plants take a long time to change output, nuclear is somewhat better than coal at load following but not quite as quick as natural gas, wind and solar absolutely can not be depended upon though, currently those power sources rely on either pumped hydro to store energy or natural gas backup (usually the less efficient simple cycle, you might get lower CO2 emissions ripping out a wind farm and upgrading the gas plants to combined cycle)).

That said, a lot of people are hoping that electric cars will be charged at night using off-peak power in which case you can just run your baseload plants all the time while having almost everything be a baseload plant (with coal or nuclear that's how you want to run them to get the best economics). Whether they will all be charged at night is another matter (I suspect that many of them will be left to charge overnight but I also suspect that people will charge them during the day at work as well), if they aren't all charged at night then you may find you need more power during the day.

Personally I'd be more inclined to do synthetic hydrocarbons with a modification of coal to oil technology (which has been proven to work) to make it a bit cleaner (say by getting the carbon out of CO2 from the atmosphere and using something other than coal burning to provide heat for the process, the second one we could probably do right now if we wanted to).

EDIT: Pumped hydro is not merely widely used but is in fact the only large scale energy storage system we have right now (and most hydroelectric dams can do it).

Using battery electric cars naturally provides the energy storage to balance the load between generation and demand http://www.udel.edu/V2G/ and also enable more renewable (variable) power generation on the grid

the energy density of a fuel is not very helpful in deciding overall system efficiency.
eg petrol has high density, but over 70% will be wasted in heat and transmission losses before it reaches the wheels, even when the engine has warmed up suitably.
compressed hydrogen will need significant energy just to keep it cold/compressed, plus the unsolved infrastucture problems of getting it into the car.
Electric motors will convert almost all the battery energy into motive force, and alow regenerative braking to recapture energy

My thoughts about these load offset strategies are as follows:

With Compressed Air Energy Storage, the idea is to exploit an existing geologic structure such as a salt dome or an abandoned salt mine (salt is good because salt has some liquid properties and hence fills in any cracks or fissures which would otherwise compromise containment). The large size of the geologic structure offsets the relatively minimal efficiencies of the concept.

However, these structures would need to be sited as close to the generation points or to a transmission line to the generation points as possible plus they would have to be near a supply of stranded natural gas since the air pressure alone is insufficient to turn a turbine. It already takes a fair bit of weather modeling and competitive bidding to secure wind generation points that are close to transmission lines as well as receive sufficiently constant winds. It's possible that there would be few geologic structures that can be exploited in this manner and most that do exist would be best exploited by natural gas generation rather than by something relatively green. Also, where salt structures exist are also areas with significant geologic instability, it wouldn't do for your CAES plant to disappear into a sink hole.

There are many substances that store a fair amount of energy in phase changes such as sodium acetate which is used in reusable heat packs but in truth all of them store less energy in the phase changes than water. That's why steam engines was one of the first technological marvels, the hydrogen bonds of water is what stores the energy and any other chemical has fewer of these H-OH bonds than water ie.: C2H5OH will hold less energy in a phase change than H2O. Of course, many such as the aforementioned supercooled sodium acetate have properties that allow the energy to be released on demand. Unfortunately, the capital costs of such a strategy is quite high as the material to hold the energy can't just be found like the geologic structures of CAES.

All of these approaches including hydroelectric are good ideas but they each have their limits either in terms of where they can be sited or in terms of the capital costs of construction. They are also all somewhat inefficient but it is energy that would otherwise just be lost and they all represent a scale much larger than is economical with chemical batteries.

Plug in hybrids are likely to be the most popular way to load shift night time generation capacity as the cost of the storage facility will be borne by the public. All the power companies need to do is provide the control mechanisms so that they can schedule the charging to optimize night time loads. Here, the high cost of the chemical batteries are not an issue because somebody else is paying for it, namely the hybrid owners.

With the concept of super capacitors and ultra-capacitors, there's likely to be major advances soon as carbon nano tubes will allow the capacitance to be exploited on an incredible scale. However, keep in mind that the voltage is linear to the charge contained by the capacitor (Q=CV) hence as soon as you start drawing current, the voltage drops linearly (with batteries the native voltage is determined by the chemistry and only drops with load due to the internal resistance of spent chemicals). This presents a problem in voltage regulation as voltage regulation usually involves some vampiric current draw and also means that the energy stored in the capacitor is unusable when the voltage drops below a certain point. Switched inductive circuits can boost the voltage levels allowing us to squeeze the last bit of energy out of the capacitors but again at the cost of some current. It's possible that capacitors and batteries could augment each other very well in hybrid vehicles.

All alternate technologies require us to redefine how we use our vehicles, the energy density of gasoline is difficult to match with any alternative vehicles hence the infrastructure has to change to provide more frequent "fill ups", such as plug ins at work, inductive coils under interstate lanes etc. Ultimately, in order to work, an incredible amount of manufactured products need to be replaced, not just the cars but the charging stations and road construction. The carbon footprint of the transition may never be amortized by the "greener" operation of the vehicles. We could easily shoot ourselves in the foot with such alternatives. This is why I advocate synthetic gasoline and diesel as the cars and infrastucture need not be replaced or modified and only the method of fuel production need to be changed and not by much.

In terms of energy density by mass versus energy density by volume, you should try to calculate the amount of mass being saved. With comparisons between hydrogen which has 39,000 WH/kg versus gasoline's 13,500 WH/kg and versus lithium ion's 250 WH/kg (sorry but batteries just can't compete on mass), the savings in mass of hydrogen over gasoline would be about 26 lbs per vehicle, this is hardly a consideration worth mentioning. Modern internal combustion engines have efficiencies far higher than 20%, the inefficiencies is from the engine not running at it's optimum rpm and the sporadic power demands from traffic patterns and lead feet, which is why hybrids are so successful. If the math is actually done, it's still energy by volume that matters most with the alternatives that we have at hand.

first off, storing electricity has loses, doesn't matter where you made it. The more power that goes on the grid, the loss to resistance goes up. Once we get more electric cars, the cost of power between day and night will go away.

storing electricity is wasteful, we should never store it except for uses in portable power needs.