A lot of chasing red herrings in this thread. The technical arguments are kind of beside the point.
People don't buy a Tesla over a Camry today because of the break-even on charging with electricity instead of gas. People buy it because they can afford it, and after that, some combination of other factors, including amenities, outward expressions of identity, curiosity, appreciation of the engineering, a huge crush on Elon Musk, etc.
The purchase of one of these home batteries at the present day and time is going to be similar.
Now, back to more of the economic/engineering point of view.
I recently talked to someone with decades of experience in electrical utilities, much of it focused on addressing peak load scenarios as part of research into my own project to try and bootstrap a business in Lithium Ion battery reuse. Long story short, there is a market for this.
The paradox here couldn't have been bigger. The grid works on the network effect: the more people using it, the more attractive it is for more people to use it.
Unfortunately, the inverse is also true - as more and more people live off the grid, the less and less attractive it is for those to remain on the grid. If living off the grid becomes wide spread (I don't think it will be for a large portion of the society), grid will fail to exist.
I'd love one of these and wouldn't want to live off the grid.
It could be used to store excess energy during the day and then power stuff at night, with the grid as the backup.
At present we can't use all the power we generate during the day.
Still it would require a massive change in how electricity is generated because the power plants which are operating now aren't designed for widely varying loads. Guess who is going to pay for that. Not using the grid is going to make other people pay through the nose for their electricity.
Keep in mind that residential use is less than half of total electricity consumption in the US, and probably other developed countries. And while some people think they want to be "off-grid" the person you are replying to (Dave) has no such plans.
You are also oversimplifying. A portion of power generation is designed for base load service. Such plants aren't designed for widely varying loads, but they tend to be designed for fuel efficiency, because they are operating at or near full load constantly. Utilities have to turn to other methods for that. Gas turbines powered with natural gas are popular for peak demand, because they can be spun up quickly. However, because their duty cycle is probably at best 20% or so, they need to be relatively cheap, and efficiency is traded for cost.
Against this reality, grid-tied solar with battery banks that someone else is paying for are going to get the attention of the power industry, if they don't have it already. Utilities would be happy to tap battery capacity during peak demand, and, in the net-metering regimes in the US, at least, they can pay it back with off-peak power. They have already dabbled in signing up institutional entities (hospitals, manufacturers, datacenters) who maintain backup generators for high-availability to provide power back to the grid in peak-load situations.
Oh, and for what it is worth, Elon Musk is chairman of the board at Solar City, which has built a nice business financing and installing residential solar. I'm sure they'll be an important sales channel for the Tesla home packs, and will be able to aggregate a lot of residential customers when negotiating with the local electric utility to provide peak demand service.