My understanding is that most of the fusion net balance problems are proportional to scale. Machinery to initiate fusion must be at least this big, consume at

My understanding is that most of the fusion net balance problems are proportional to scale. Machinery to initiate fusion must be at least this big, consume at least this much power, and cannot be scaled down arbitrarily. However, if one were to scale it up in size, the same doesn't hold. Output power scales faster than increased input requirements. Consequently, most current fusion work is (a) find a design that theoretically has those scaling characteristics, (b) build a prototype to investigate / prove any unknowns (net negative power, but not ITER/NIF expensive), & (c) if able to prove (b) then scale up into a net positive example.