The output of new panels hasn't risen much above 14% to 19% efficiency. There are a bunch of exotic chemistries, but they are toxic, expensive, and only really used in applications like space probes and the space station. Many of the solar plant closures are happening well before the design lifetime of the plant. It's even worse when you consider the relatively poor EROI of solar panels. These things need to run as long as possible to amortize their embodied energy.

This is a trend I don't quite understand. We have known for a long time that old solar panels continue to produce at most of their rated capacity for decades. But at the same time, I see a huge glut of used solar panels coming mostly from the closure of utility scale solar power plants built in the last 15 to 20 years. I mean, it's GREAT for people who want to install DIY home solar. But are utility scale inverters and other consumables so expensive that it is cheaper to scrap the entire plant after 15 to 20 years of operation?

We do in fact have refrigeration systems that use gases as the working fluid. They utilized the Joule-Thomson effect (most gases cool upon expansion) and were called "Bell-Coleman machines" or "air cycle refrigeration machines". And they do exactly what you suggest, use a turboexpander as an energy recovery device rather than an expansion valve, the throttling losses in a gas system being too deleterious on performance to ignore. Even with this, the refrigeration COP was rarely greater than 1. This method is still used in air liquefaction devices in air separation plants, and for aircraft air conditioning where compressed air is readily available by bleeding from jet engine compressors. The low efficiency and the large circulating volume of gas needed per unit of refrigeration put them at a distinct disadvantage compared to mechanical vapor compression refrigeration. Here is a guy who built an air-cycle air conditioner powered by a vacuum cleaner and with a turbocharger as the turboexpander.

https://www.youtube.com/watch?v=f1FQjfyOifI

Many of the impoverished people now gaining access to electricity previously cooked their food by burning wood and animal dung inside the dwelling, causing millions of air pollution deaths. For them, coal-fired electricity generation may be a significant improvement in their health. Also, a lot of the air pollution in India comes from the open burning of rice stubble.

We don't have fusion. Even if it is viable, it won't be commercialized for decades. Nuclear fission is the lowest risk energy technology, period. It is the safest energy technology in existence.

Yes! I was just talking about this earlier. Natural gas utilities are bankrolling a massive anti-nuclear propaganda campaign in media.

The reason we don't do it is because it doesn't currently make economic sense. The unsubsidized price is related to its energy intensity, so it likely doesn't make energy sense to do it. It does, however, make lots of sense to incinerate municipal solid waste in waste-to-energy plants and recover the energy as electricity and steam for space-heating and cooling.

This simply isn't how reservoir economics works. At the start of industrial revolution, we had access to shallow reserves of high-BTU thermal and metallurgical coal, and geopressured oil and gas in highly permeable salt domes. Those days are gone. The quality of our coal, oil, and gas reserves has been declining. For example, people say "peak coal" in the USA was in 2007, but the peak thermal content of the coal was in 1996. (Also, keep in mind that coal only peaked in the USA due to an explosion of fuel-switching to natural gas). As the quality declines, the price increases, and this drives investment into much larger, but lower quality reserves. So now we are using sub-bituminous coal and lignite for power generation, and hydraulic fracturing, horizontal drill string steering, reservoir simulation, and deep water drilling for oil and gas, fracking and horizontal drilling being especially important in impermeable shale formations. As the price rises further, it would make other technologies viable. In electricity, space heating, and possibly industrial process heat, I could see nuclear fission replacing fossil fuels. Conventional nuclear fuel reserves could last 10,000 years, and seawater uranium could make that hundreds of millions of years. Liquid oil has extreme portability which makes it uniquely useful for transportation and heavy industrial equipment, which also makes it extremely difficult to see what could replace it. High prices on oil will eventually drive the conversion of coal and natural gas to liquid fuels using the Fischer-Tropsch process. The reality is that the fossil fuels could probably last a few centuries. The declining quality of the fossil fuel reserves also means that they are becoming more and more energy-intensive to extract, which also leads to declining Energy-Return-On-Energy-Invested, which has its own set of implications for industrial civilization.

No, transportation and mobile industrial equipment are why oil, compared to coal or natural gas, is uniquely valuable. There is no replacement for it, especially in heavy transportation, freight, aviation, maritime, and mobile industrial equipment.

The catastrophic visions keep coming, and yet the standard of living and human lifespan keep rising, especially in the developing word. The recent environmental campaigns titled "keep the X (coal, oil, and most recently, natural gas) in the ground" are an admission that technology has increased our recoverable fuel reserves so much that the original predictions of resource depletion is now now longer a concern for the foreseeable future. And whenever people want to get serious about clean energy development, billions of tons of Uranium and Thorium lay waiting. The few things that genuinely could hamper the rate of technological progress are inflation and precipitous energy cost increases, both of which raise the cost of research and development, and production of high tech devices. But the current price increases aren't related to resource depletion.

I think a good parameter for the (accelerating?) rate of technological progress would be the number of new drugs making it to market produced or theorized using AI drug screening and prediction capabilities. This is something that has the potential to massively increase human quality of life and lifespan and it is already beginning.