Well, there are thousands of interesting micropython projects out there, hardly any of them appear forced.

I needed a custom motor controller that I can talk to over a webserver. It performs closed loop control via an encoder, controls DC auxillary motors and now ran without hiccups for years.
A "real" solution was estimated by an engineer to take four weeks. Micropython allowed to iterate on this in the field very quickly, and within a few days the system was completed.
Could a smaller microcontroller do the same thing if I had used C? Probably. But unless this would go into tens of thousands of units my time is more valuable than the cost savings going from $1 microcontroller to something cheaper.

If we assume a best case scenario, that is all sunlight is captured by the 60 x 60 feet reflector and then send down to earth in a 3 mile diameter circle this would correspond to a light intensity of approximately 0.02 W / m2 or 2 Lux.

This is barely brighter than the light from a full moon. Probably not even enough for any color vision. So in which scenario does that help? And that already entails that a full satellite is only dedicated to you. Someone with more economical knowledge than me might want to give an estimate what the hourly rate of a satellite of that size might be.

The whole idea then goes brr by assuming thousands of satellites (1000 Lux would be bright office lighting) which is still not enough for any photovoltaic usage. So this is only an investment vehicle for people that dream without basic math.

If we would leave the monitors in the ocean it would be possible to turn them back on.

This is evidently not an effort to cut costs but to spend money on dismantling a system that provides inconvenient information. Once dismantled it will be significantly more expensive in any future scenario to get this system back online. Probably now even more than the initial $368 million.

Exactly, and this goes hand in hand with deterministic behavior.

I can test and tune a microcontroller-based system and after that can be certain that it will behave almost identical if deployed in a remote location, inside a larger machine, etc. With a linux-based system there are many more moving parts, some driver can run several threads in the background, on x86-hardware there are the NMIs that can cause a lot of jitter, etc.

Most likely there will be pressure from both sides, a Teensy microcontroller is almost as powerful as the early Raspberry Pi systems. On the other hand a Pi Zero is less than $10. So there is quite a bit of overlap already but running a full OS like Linux does not feel like the solution to every application.

Well, in Alaska or the artic region these panels would work even worse.

It is a thermoelectric generator, so it requires a Delta T to operate. Here the cold side is the open sky, the warm side is the panel and ground underneath. If you reduce this temperature difference by going into a cold climate the efficiency drops rapidly. It might provide a slightly higher benefit in the Sahara desert. But as others here calculated even a tiny battery would provide more power than this device.

CE means that the product can be sold on the European market because it complies with basic rules and regulations. And yes, the manufacturer or import company is responsible and an external certification is not required.

There are different rules though for "ear lights" and medical software. The ear lights are in the toy category, so it does not need to fulfill any function, so the regulatory bar is also very low. Anything that claims medical benefits has many more hoops to jump through. So if someone markets or advertises a toy as something that cures with depression is lying but I am certain their CE mark (and you can ask for the regulatory documents) do not state any health related benefits.

This might help in applications where battery storage is hard to implement.

Still, a good solar panel averages at 150 W/m2. So this provides only 0.0003 times the output power during the day. And thermoelectric materials don't come for free, so I don't know if this will ever break even with the energy that was necessary to make the thermoelectric materials in the first place.

That is an interesting question but consider the evolutionary scenario where you might enter a room with concentrated N2.

This was almost impossible until the last centuries and so therefore there is zero evolutionary incentive to develop a response to that. On the other hand scenarios that include too much CO2 are abundant, from holding your breath for too long, a small cave, and many other situations lead to an excess of CO2. And additionally if your lungs don't fully work you can compensate that to some extend by breathing more rapidly. In none of these situations a chemoreceptor reflex to N2 would be particularly helpful.

Furthermore it is easier to have a "sensor" to detect small levels of something than to detect slight changes from 78% N2 to 85% N2. That would require fine tuning.

Not only is credit often given way too late, the cult around specific persons does not reflect at all modern scientific practice. Just look at the article:

About 20 scientists working hard for years on a project together based on a 15 year old idea. And the 50 citations won't even cover all the necessary requirements to get this off the ground. And with an increase in complexity this trajectory spans beyond science.

230V * 16A = 3680W, so 3.5kW, not megawatts. A megawatt cable requires a cross section of more than 1000mm^2, the conductor itself then needs to be about the thickness of a large finger.

So yes, still doable but we are now talking easily about more than 20kg per meter, so not something you can easily drag to your truck.

Yes, absolutely. The thing is that a rail or tracking system that can support large parabolic mirrors under potentially strong loads from wind might be the most expensive component of the entire system. Not that it can't be done but compare this with a simple solar panel mounting rack without any movable parts.

There is a niche for something like this but in the large majority of cases you want to avoid any system that needs to track the sun.

Well, a homeowner can heat water during the day or generate electricity with the steam using a turbine, a thermoelectric generator, etc.

For industrial use you want steam available 24/7, so this can provide additional energy but definitely won't be the main source. Unless you combine this with molten salt storage but than you basically replicated the Solana Generating Station on a small scale with all benefits and issues.

It is not snakeoil. There is not really a physical limitation why such a device cannot reach 85% efficiency.

But hold on with the yay just yet. This is based on a parabolic mirror. So it gives you electricity and steam under two conditions: The sun is shining (cloud cover reduces both visible and infrared components a lot) and you need a mechanical system to move the mirror. So instead of the more or less simple installation of solar panels on your rooftop you need a large dish with a tracking system in your garden. Scaling up means probably multiple mirrors each with their own tracking system. Feasible but expensive and requires maintenance.