Refrigerators/freezers at home is the least of our problems.
Assuming nobody is opening them during the nights, which is a fair assumption given most people would be sleeping most of the night. A modern refrigernator/freezer would probably run the compressor for just a few minutes during the whole night as the isolation is fairly robust. I'd be surprized if they use more electricity thorugh the night than a single 60W light bulb or maybe 2 during the same period of time. During the day it probably is more.
We have far more stuff at home that will put load. In particular air-conditioning units. As it happens, the places where solar panel makes more sense, and during the time of the years that the sun is going to provide the most energy also makes it likely that the people living there will want to have air condiotioning. This alone will make the fridge's energy consumption pale in comparison.
Solar panels without any sort of industry-changing battery system would not be sustainable. Traditional batteries are unlikely to be helpful in a large scale. For one, their energy density (volume and weight) are still quite poor and they are very exensive. The materials used to create these are also going to be damaging the environment. Also remember, if you were to transition to 100% renewable with solar/wind as your bulk, you would need large amount of redundancy for energy production and large amount of redundancy for energy storage as to account for time of maintenance and off-time for nighttime, cloudy days (or weeks depending on geography), etc. So you don't need to cover 100% energy production, you need to cover 200 or 300% + 200-300%+ storage.
Other "batteries" such as molten salts and/or water pumping might be helpful for larger scale but I still think you will need to supplement renewable energy with additional sources.
We also have the issue of cities. In suburbia and farmlands you can stick a whole bunch of panels on the properties' roough and that is area that is hit by sun and not being used for anything else (in particular trees/farm). But you cannot reasonable do the same on a tall building. For instance, if you had a modest 7-10 floor building and you cover 100% of that building's roof, you are probably be hard pressed to cover enough energy to run the hallways, elevators and perhaps one of the floors.
Energy efficient devices + on-site battery solutions does not help with energy requirements for public "assets". For instance traffic lights... you need to run those as well.
In the unlikely event that solar were to become the major source of energy, we would need to outproduce our needs by a large amount. We would need significant redundancy in production as to account for rainy days, cloudy days, dust storms, summer peaks as well as our general growth. We could have a scenario where we end up clearing large amounts of terrain on sun-rich areas for this task. The flipside, is that those areas are also the areas that would be growing plants, trees that contribute to CO2 sequestration from the atmosphere.
I think nuclear is the best way forward. It got bad reputation but compared to hydrocarbons its not even playing in the same field as far as safety is concerned. All major accidents of nuclear combined amount for a tiny fraction of people that have died as a result of hydrocarbon energy production. That is not without taking into account global warming as we have no idea the end-result impact of this.
The current generation of Nuclear power, are vastly superior to those that had the major accidents that we remember in "collective memory". A few of the accidents actually would not have occured if the plant's automatic safety measures would have been allowed to proceed and/or humans followed guidelines, in other words, our stupidity got in the way. That being said, newer designs make it extremily unlikely for these accidents to happen. It wouldn't be 100% safe, but it would be pretty damn close, certainly far closer than just pumping poisonous gas into the air we breathe.
Nuclear comes in 2 flavors and the one flavor we have access right now does have a weakness, the byproduct of radioactive waste. Since the topic seems to be around ACC/AGW this does not need to be discussed at length, however suffice it to say that there are methods to minimize this kind of waste, there are technologies that could completely eliminate the "worse" type of waste (the ones that have half-life in the thousands of years) such as thorium-based and/or reactors that can use non-enriched fuels and/or re-cycle waste.
The holy grail ofcourse is fusion energy which is clean, and in theory at least potentially far more efficient than anything that we have come accross (in fact other than using blackholes and/or anti-matter, as far as we know this might be the most efficient way to convert mass into energy). Fusion is something that we have done for decades, so it is possible. The big issue was that for many years we needed to put more engergy in than what we got back, this is no longer the problem. Now the focus is in trying to make this last more than a few seconds as the "successful attemps" have lasted in the orders of seconds or a few minutes. That being said, the consensus is that we should have this technology up and running in the next few decades.
I think we should embrace a combination of nuclear (fission) and renewables (solar/hydro/wind/etc) for the forseable future. Once fusion becomes economically feasable we will have a lot of cheap clean energy, enough that we could afford to literally take carbon off the atmosphere and reduce it to our will. How could we do this? well we could grow trees/plant organic matter to sequester the CO2. For instance, if we grow a few tons of lumber, and use that lumber to build large buildings, the CO2 inside the lumber would be "permanently" sequestered from the atmosphere (as long as we don't burn that building down).
With sufficient cheap energy, we could use areas of the world that are not "productive" during times of the year (higher latitude winter season) and artificially introduce energy in the form of light (large LEDs) and literally force growth and carbon sequestration. We could introduce light into parts of the ocean which have the potential (but not the light) of being productive, we could potentially growth billions of tons of plakton in parts of the ocean which cannot do that because light does not reach (remember, despite the ocean cover 2/3 of the surface of the planet, it is the coastal reefs that actually have the most productivity in the seas). We could artificially change desert landscape by taking salt water and making it fresh.
Our best option is to advance our technology, and to obtain technologies that makes it cheap and possible to have a significant impact in our planet at will (as opposed than as a by-product). Cheap, abundant energy could allow us to literally terraform our own planet at will, that is what we need. I don't think we can do that with traditional solar. I believe we will need nuclear.
Eventually, centuries from now, or even millenia if humanity is still around, we might get "solar" in the form of huge arrays of solar collectors and basically form our very own Dyson Swarm. But that is too far away to think about it.