The efficiency comes after a critical mass has been installed. Right now, energy storage (battery installations) have still not achieved critical mass in California, whereas in South Australia, they have. It is just a matter of volume and cost, and both of those are being hindered by resource cost pathways that are disallowed by regulation. When the threesome of regulation, investment, and supply have been corrected, California can achieve the critical mass of energy storage (battery installations) necessary to gain the efficiency needed to make them cleaner. This is a process, not a single chicken we're cooking; you achieve one chicken cooked and ate after killing one chicken, but you don't achieve saved energy for 30,000,000 modern people just because a few thousand PowerWalls and a handful of PowerPacks were installed. There are a lot of interconnected parts, and we have to go interconnect them for it to function.
That's going to be proven again and again. Stick in the muds will look at the latent numbers that poop out the end of each chicken without realizing we're talking about whole farms of chickens, to keep up with my chicken metaphore. This will get really big before we can look at efficiency numbers. We already get an inkling of what is possible in installed systems from the South Australia system, but even that hasn't reached full critical mass yet.
Since Tesla is battery starved for its cars already, it will be a long time before there is enough energy storage supply capacity to reach that critical mass in California. It will be quite some time yet before we can measure current pollution levels as being indicative of the efficiency of this system overall. It really is conceptually as simple as sunlight to solar panels to batteries to users. It is just a matter of getting those batteries out there.
By the way, many of the energy storage devices are already in Tesla cars and other electric cars. That means that the solar should charge the cars. Another investment that must be done to reach the critical mass of solar clean energy is charging direct from solar panels to cars in parking lots at work: that is where the car is parked when the sun is shining, so that is where it must be charged to maximize efficiency; in that case, the energy storage is built in to the car. Once again, regulation needs to get out of the way so that resource cost pathways can connect the car owners and drivers to the employment parking lot solar canopies and plugs that would plug into their electric cars to charge them. While I think welfare is an inefficient buffer for establishing resource cost pathways, one concept for SGIP could be to plug SGIP into cars that charge at work, i.e., fund any missing part of that, such as work parking lot solar canopies with charge plugs, which would compensate to whomever owns that solar canopy; regulation would have to allow it to be spurred by electric car owners and drivers, so perhaps the car drivers could own a share of said canopies and get the SGIP welfare encouragement resource allocation cost pathway methodogy; after all, it is another form of energy storage, as I pointed out. Have the chicken eater buy a share of the chicken coop. Or something. But don't expect a single chicken to tell the whole story. It will just say "bock bock bock" and poop.