> Yes, any new thing is expensive. These points are not necessarily intrinsic to the process.
It's not expensive just because it's new, it's expensive because it's trying to do a very very difficult thing - using magnets to achieve what the entire mass of Jupiter can't achieve, compress hydrogen so much that it starts fusing, and then keeping it compressed while it's essentially violently exploding - and exploding in a rain of extremely fast heavy particles that don't interact with the magnets at all.
Yes, the steel required to withstand the force of the magnets, and to be dense enough to prevent hydrogen from leaking, magnets powerful enough to contain thebl fusion reaction, cooling systems to keep the superconducting magnets in close proximity to the neutron rain at extreme low temperatures.
These are all the parts we know about. Then, there are all the systems that no one has attempted yet that you will need to actually extract some energy from the whole thing, and to inject fuel into the running reactor, and to recycle tritium.
Overall the reactor vessel has to be built similarly to a high-pressure submarine, but it needs to withstand even higher forces. Not exactly something that can be done cheaply, even though we have been building submarines for a good 50 years.
We know the energy density is there, based on thermonuclear weapons.
Yes, the designs for a power plant that are similarly impressive don't exist today. That's where research and engineering can help.
Yes, any new thing is expensive. These points are not necessarily intrinsic to the process.