This is usually less than the efficiency drop involved in a round trip through a battery. It's not as big a factor as you think.
This seems to me the big benefit here -- enough penetration of home backup batteries means public power generation doesn't have to be built up to provide massive peak surges. Further, trickle charging the backup batteries during the evenings/nighttime and allowing them to meet some needs during the day also means a higher net usage of generated power, so perhaps levels of generation could go down altogether.
There's also then the capitalistic angle -- if you want more peak power at your house, you can buy more batteries in lieu of (or in concert with) installing secondary power systems.
But reducing gross energy production due to higher net utilization would be a really cool thing, so long as the total cost to the macro system (considering full life of the battery vs. load taken off power plants) is a net positive.
Electricity, on the other hand, has no viable localized option today. It's also the only one of these that has a significant drop in efficiency due to the distribution itself and has a significant impact to the environment.
Local power storage means that grid can balance the load between peak and non-peak times. Also, local power storage means that wind/solar now has a solution for time where power output is reduced.
Having a battery that can power a home for a week is huge, if it's affordable. This could significantly reduce power generation costs.