This is an incredibly naive bordering on "probably just dishonesty masquerading as ignorance" thing to say.
Of course cost goes up fairly linearly with size. More rebar, more pipes, more materials, more diesel in the bulldozer, more man hours, more welding rods, more welds to inspect, more money.
But there are also a ton of one per instance overhead paper pushing costs, especially with something as onerously regulated as nuclear that are a large fraction of overall costs and will be roughly the same whether you're building a 1kw facility or a 1gw facility. All of that has to be amortized over your reactor. So if you build a small reactor each dollar of reactor might have $10 of overhead it needs to amortize. If you build a big one each dollar of reactor might only have another $1 it needs to amortize.
Of course there's a sweet spot before the reactor is so massive the part count grows so high that the MBTF starts getting you. That sweet spot also has a low end below which the fixed regulatory and palm greasing costs are so high that your tiny reactor can never hope to pay them back in it's lifetime.
The argument being made doesn't require a linear relationship between size and cost. Remember, we're looking at two objects with the SAME power output, one much larger than the other.
The MTBF/MTTR of fusion is actually looking like an extremely serious problem. Fission reactors can operate with 2% of their fuel rods leaking. A single leak through the vacuum boundary of a fusion reactor will contaminate the plasma and shut things down. And repairs on a fusion reactor are going to be much more difficult than simply replacing fuel rods. The innards of any fusion reactor will be so radioactive that hands-on maintenance will be impossible.
Of course cost goes up fairly linearly with size. More rebar, more pipes, more materials, more diesel in the bulldozer, more man hours, more welding rods, more welds to inspect, more money.
But there are also a ton of one per instance overhead paper pushing costs, especially with something as onerously regulated as nuclear that are a large fraction of overall costs and will be roughly the same whether you're building a 1kw facility or a 1gw facility. All of that has to be amortized over your reactor. So if you build a small reactor each dollar of reactor might have $10 of overhead it needs to amortize. If you build a big one each dollar of reactor might only have another $1 it needs to amortize.
Of course there's a sweet spot before the reactor is so massive the part count grows so high that the MBTF starts getting you. That sweet spot also has a low end below which the fixed regulatory and palm greasing costs are so high that your tiny reactor can never hope to pay them back in it's lifetime.