The general thought is that our increased demand for electricity will make up for the energy production savings generated through other "private" sources. - That said, many years ago we volunteered for a time of day, day of week usage scale of billing. We, along with what apparently were a whole host of others in the area, signed up for a minimum of 1 year commitment although the program was ended within 6 months for the very reasons you state. For ourselves, someone in the house set the washer off at midnight and, someone else got up an hour early to put it in the drier for a 75% plus savings on the cost of the electricity used. To capture the cheapest time slot of mid-week 12M-5AM, if we were home on Friday night, we would take a nap then have a Friday night electricity usage party. Activities using electricity like major laundry, baking, cooking, dish washer, vacuuming, and even tools in the garage would commence at what we counted down as 12:03 just to make sure. After the program was prematurely disbanded, people began to quietly grumble and discuss the electrical usage schedule changes they made for close to two years after; we were not the only folks who actively changed their schedules, used more volume of electricity than they had before, and still paid a lower bill.
.....I am not as sure how thin some of the margins truly are; at least in the past once the power brokerages were broken. Also, a number of the companies/systems have pulled in their profits rather than doing maintenance working for performance upgrades. For many, there is a lot of ready ROI available to them for simply maintaining and upgrading their existing systems. For example, most households that transitioned/are transitioning from incandescent light bulbs to either CFL or LED did not turnkey their entire household all at once; rather they added a few here and there due to the conversion costs. Whether they were aware of it or not, their electrical savings probably started to be enough to cover the cost of additional bulbs. Likewise, people needing to save money and switching to power strips so they can completely shut down electrical items seldom do a whole house conversion at once. By concentrating on one or two and their highest sleep-mode power users, then reinvesting the money saved on electricity they can effectively pay for all but the initial purchases in saved utility costs.
Most local electrical grids needlessly loose electricity. Some loose more than 50% of the electricity they pay for and then pass on in utility costs. Much of it is due to the wires, connections, and switching systems they use. On a greater scale, one heck of a lot is lost as power flows from one local grid system to another for a whole host of reasons. Generating more efficiency should help them recoup a lot of their generation and transmission losses.
In theory, if a local company generates their own power then, they should be able to realize a profit on reduced generation costs and/or by selling their excess. Likewise, if they can absorb any excess generated for them using private generation sources within their grid, they should have less generation costs, surplus to sell, and/or less external electricity to purchase from outside their grid. Reality is, units of electricity are sold up to a year or more before it is actually used. This lets a company lock into a generation/purchase/sale rate thus better enabling budgets to be made and, theoretically saving money. Not unlike agriculture, on a good year one company will be able to meet their own needs while selling excess out on the open market while some other company has so much need that they are willing to pay a premium price. Of course, on a bad year a company will be out scrounging for what is available at what ever is the going rate. If I remember correctly, payment is/was made on the amount of electricity sent out onto the grid rather than what is actually received by the buyer; another argument for getting the grid a little more efficient. There can become a time however, that between what is available for purchase, what a company is able or willing to pay for surplus electricity, what the customers are actually using, and what is lost in transmission there just plain isn't enough electricity. And, this is what happened during the California, rolling brown and black outs.
I will guess that as more electric vehicles, 50-60 inch wide screen TVs, computing cloud storage sites, and even domestic manufacturing come on board, our collective demand for electricity will increase rather than decrease. If, when, as we shut down coal production permanently and/or emission or efficiency upgrades there will be plenty of potential advantages to these little, private electricity generation sources. All the more so if/when/as utility companies find fiscal advantages and/or are mandated by customer demand and/or regulated to utilization of these micro power generation plants.
Among other things, what happened to our local utility company was that they effectively generated more prime time and over-all usage savings than they thought possible; which was FAR less than they estimated as probable. There was an up front and back room, loud and quiet, hue and cry from the M-F, 9-5 business community on the home ranges. The sources from which they purchased power would not cooperate in re-directing prime time electricity nor assist in the selling of the excesses. And, the electricity brokerages were in full operation at the time were not on board; they certainly didn't take kindly to the local utility nor any others in discussion for the use of the excess, prime-time electricity. That said, there have been periodic whisperings of adding prime time surcharges locally. Basically the same time of day, day of week structure as the rate savings program of yore but, instead of saving up to 75% for the weekday 12M-5AM (or 1AM-4AM) that time slot will be for the basic rate; the other time slots will have a series of increasing convenience time slot surcharges. Applicable to residential accounts only of course.