Calling what the 19 year old Taylor Wilson built at age 14 a "fusion plant" is somewhat misleading. For me, a "fusion plant" is a device that generates power. His machine, a Farnsworth fusor, should more appropriately be called a "reactor", because it can only sustain a fusion reaction, but it cannot generate power. Building Farnsworth fusors is not too hard, given access to the right tools.
Building fusion reactors that can be used as power plants, on the other hand, is an entirely different story. The design is a lot more complex, because plasmas don't scale well and tend to get unstable as they get bigger.
For comparison:
A Farnsworth fusor can be built for around 1000 Euros, but it's Q-value (=output_power/input_power) is so low that I have trouble even estimating it. Based on a quote of "multiple hundreds of thousands of neutrons per seconds" for describing a "dangerous design", the fusion power is somewhere between a nanowatt and a microwatt, which makes the Q-value something like 10^-10.
For more complex fusion reactor designs based on magnetic confinement, the best Q-value that has been achieved so far was a value of 0.7 at the JET facility in the UK, 1997. A new reactor is currently being built in southern France. It is called "ITER", and it is hoped that it can reach a Q-value of around 10. This is the type of Q-value that you need for a fusion power plant. ITER is projected to cost around 16 billion euros.