That is assuming that there is an elastic supply of needed materials. There may be bottlenecks in supplies that drive costs up.

Is it easier to keep a city of 10k or 100k disease free? If both get "contaminated", which is the bigger cost?

That's a false analogy. 10 separate factories in their example would be 10 separate cities each with 10k people, not 1 large city with 100k people.

That's not how you reach economies of scale.

We all know that. Their point was that you get linear scaling using that approach, which disproves the notion that you can't get linear scaling.

Linear scaling of any project large enough to disrupt an industry (as is being suggested here) is a pipe dream.

When you are trying to scale up something that big, you eventually get beyond economies of scale for the required infrastructure and feedstock, and as a result, increase overall demand, which raises prices.

You only get near-linear scaling for a very short, initial "ramp-up" period.

Beyond that, you start cornering markets, which makes prices spike, not decrease or even hold steady.

>> A ten times larger facility would cost more than ten times as much to run.

> This is an impossibility. You can just build ten smaller factories to scale up and get at least a linear gain.

No it isn't. Your "solution" is a different thing. A collection of smaller facilities is not the same thing as a larger facility.

You can have a large facility with 10,000 automated lines.

> You can have a large facility with 10,000 automated lines.

Read the article. IIRC, it's talking about the limits of how large you can practically scale a clean room.

Environmentalists often have a bias for small scale operation. They just really like the idea of everyone having their own bioreactor in their closet powered by their own rooftop solar panels. I think it may be simply an aesthetic preference that they then try to rationalize, facts be damned.

It depends, if they share some facilities with other functions that pay parts of the bills then those parts wont scale up. Like lets say you rent out part of your capacity to a nearby lab that pay research lab money for it, that could cover most of their expenses for a small facility. I wouldn't be surprised if that is how they are running things right now.