If I were picking a number out of thin air I'd certainly go higher. The question is how he arrived at that number, and I suspect economies of scale have something to do with it. Scale can do funny things to your calculations. Things can get harder, then easier, then harder again as you go up.
I once had a colleague whose first engineering job out of college was to do a reverse engineering specification on a prototype submersible; the Navy was pleased with the low cost of the prototype and thought it might like a second one. The problem was that the prototype was made by scrounging surplus bits and pieces; building a second one that was exactly the same would have cost 10x as much because you'd have to hunt down an exact replacement for each part. But if you were building a 100 units, you probably could do it for 100x the cost, because you could amortize the project's fixed costs over more units: source the odd parts, reverse engineering them if necessary, or doing design revisions with an eye to making it reproducible.