I was talking about the most efficient way to use a generator where you have to have a generator, and that would be off-grid or where UPS is needed.
On the 900MHz thing, it was a point to point tower. So they did not need to run fiber if they could get power. I think if they were running fiber there we'd have though of supplying power lines too.
On car inverters I am not talking about the _alternator_ or the regulator (which just rectifies the output from the variable frequency alternator, which is not 60/50Hz wall current, into DC at around 14V) - I was talking about cheap inverters that plug into the lighter socket on your car and provide 120V 60Hz (in the US) alternating current (AC) back for running crap like a laptop. Inverters are typically not built into cars.... I didn't think that would be unclear (since it was I'm not sure why I'm continuing to write the rest of this since you don't understand any of this stuff). The cheap inverters you buy having cheap switching components.
Lead acid batteries are not being considered in any serious way as an industrial storage mechanism. Pump hydro already exists, and we're trying out flow batteries (nasty chemicals but a close industrial loop), sodium sulfide cells (we've got a 6MWh storage system at a bus station in Long Island for time shifting), and even lithium ion if it can get an order of magnitude cheaper.
The reason power is cheap at night is because of technical limitations with our generation systems - Niagara runs because if it doesn't the water is gone. So for Niagara they build Lewiston pump hydro which allows that power to be load shifted to the next day. Most steam plants have a need to be started up and shut down gradually because of torsion on their shafts that go from multiple heat stages to the generator, so idling them entirely overnight is not practical. Nuke plants have the steam plant issues plus criticality requirements. All of these cause us to generate power at night which is what causes the cost differential. For generation that can be started or stopped quickly - gas fired turbines - this is not true. As more generation is made gas turbine the differential will disappear. This is actually an economic problem for the facility I work at. Anyway...
As for "safety"... I work in a 345kV power facility. I work with 500kW paralleling diesel e-gens for black start, in stators that run at 16.9 kV for 325MW power production, in a switchyard where 345kV lines are about 25 feet off the ground... I've also got solar at home. It is safer, if by safer you mean human safety... I don't have an arc flash distance of many feet on my 96VDC power. If you mean dependable or dispatchable I might go along with that.
You might be having a cognitive failure about what gives "redundancy". If you have a system that is 100% dispatched although you may have many many generators none of them is redundant. As it stands now on the grid (at least the part I know which is New York state) we have 18% excess capacity ready to go at any time, guessing what the power demand will be. However, the generation is all over the state, so there is the separate question of do we have excess transmission capacity, and at the moment that is usually barely enough for NYC metro. So a possible solution to the bottleneck is more local generation inside the NYC area. We're looking to do that with distributed generation (DG) - emergency generators at telecom central offices, other utilities, businesses with UPS, and yes, solar power. (There's also a plan for a 500kVDC cable from Albany to NYC under the Hudson... we currently run a 345kV underwater from one side of the Hudson to the other, and based on that I'd say the DG solution is... a tad less complex.)
As for "bad for the enviroment"... well, I'd say you really just don't want to know how the sausage is made... so you don't want to know about the underground PCB oil filled cables that still exist in many urban areas (because there is no viable alternative) or the SF6 gas breakers that we use, that can theoretically decompose into SF10 (a neurotoxin similar to phosgene) and is 23000 times more potent as a greenhouse gas than C02 (again, no viable alternative so we've got a waiver) or the lead, mercury and other crap you're currently getting from the coal plants... jeez, what stupid argument that is. Assuming solar has to be made as it is now from silicon wafers (which is not necessarily true for solar PV and has nothing to do with with concentrating solar), it is possible to do that more cleanly than burning coal... I worked in a chip fab in New Jersey that did gallium arsenide chip fab and we recovered all of that nasty crap like chloroform from the water and recycled it. Easy to do in a manufacturing facility, nearly impossible when you're in a combustion facility because of the amount of mass flow you're dealing with.
Btw, everything fails. Things with no moving parts tend to fail less. Where heating and cooling cycling make things that aren't supposed to move move they do fail more, but all things being equal the solid state devices work a lot better than their electo mech equivalents... which is why we can do things with switching components that we could never do before in motors, generators, statcoms, cap banks... so if you've got a solar facility, things are failing continuously, as they are at every power facility in the world, all of the time - but the failures are insignificant, unlike say when we have a water blockage from a broken flow switch paddle cause a heat exchanger to fail to cool oil on a bearing in a 325MW stator.... Or any one of the literally thousands of other points of failure from that 325MW generator - cooling systems, wicket gates, spherical valves that have thousands of valves themselves, electrical control systems, breakers... and I'm talking just a hydro generator which is dead simple compared to a coal plant, which in turn is far simpler than a nuke plant. If you are saying solar of any type has similar complexity you simply don't know what you're talking about. But that was clear.