Torvalds just really dislikes C++. He’s gone on the record saying that he thinks it’s just not a good language. In his own words “C++ is just a waste, there is no design at all, just adding some scum on top of C.”

There’s a whole bunch of mechanisms, largely depending on the fusion architecture and the atoms being fused. For tokamak reactors the circular nature lends itself well to what you describe, though usually it’s energy being imparted into the ions to keep them contained and away from the walls. In the ‘standard’ deuterium-tritium fusion model (the easiest to perform) fusion produces a helium nucleus and a neutron, where the neutron gets most of the energy. Since a neutron can’t be contained by magnets it impacts the chamber walls. This heat is wicked away by, you guessed it, cooling water which turns into steam. In order to use a direct energy conversion strategy you need a fusion reaction that produces no neutrons, but we’re not there yet.

Some types of fusion can bypass steam generation and use what’s creatively called Direct Energy Conversion. If the fusion products are charged particles they can be passed through a magnetic field to separate them based on charge and collected onto plates. When you look at the electric potential between the plates you’ve effectively created a voltage, no steam necessary. It’s also theoretically possible to do the same with some types of fission products too.

Because planets are terribly inefficient? Their livable area per mass is very low. You could deconstruct a planet and use the materials to make O’Neill cylinders (or similar) that can comfortably hold many orders of magnitude more people. It’s also very energy expensive to go up and down the gravity well of a planet, increasing the cost of all space-related industry, shipping and transport. For a interplanetary civilisation that’s a huge cost. Artificial habitats have no such problem.

It’s interesting the article focuses so much on Dyson spheres considering they’re probably the least practical megastructure. Dyson swarms are a much more practical and accepted concept, hell it’s one of the things that astrophysicists are actually looking for out there.