Buying and selling used goods is absolutely not that. You're talking about searching, inspecting, paying (cash), storing inventory, selling, shipping, etc.
And, no, eBay and Craigslist are not examples. They don't buy things. They are a marketplace and that's not what the original post set out to do.
Selling burrito franchises, car dealerships, and grocery stores all work very very well at scale, and are very scalable.
So Amazon was never a scalable business? Or netflix? You don't have to be sitting around a kitchen table selling nothing but bits for it to be scalable business.
Amazon wasn't profitable for a long time and only scalable because of massive investments. They don't buy used goods and resell them. Same for netflix.
His main argument is that the acquisition of items for resale is not scalable, because you are dependent on people to supply used goods. You cannot make the market provide more used goods to meet demands. Amazon and netflix can perfectly well do that.
My point is that it's folly to try and argue this particular negative. I'm not so confident it's impossible to scale reselling of used goods when there's been so much growth in optimizing operations processes.
Secondly, "inspecting" may boil down to a similar process of QA, but I think it should be obvious that assuring quality in a production process of new goods is a very straightforward thing while for used goods, you probably need a new process with every supplier you find.
And I agree that there is an astonishing growth in optimizing any kind of operations process. However, that still doesn't increase supply. This cuts back to his main argument: "grow revenue without costs growing proportionally". If you have a finite supply of used Aeron chairs, acquisition of the remaining chairs becomes increasingly hard - up until the point where there aren't any, anymore.
So yes, I added the argument of finite (if only temporary) supply, but it was only to extend his argument of proportional growth of cost. And yes, there is room for optimization and that may be sufficient to keep a business afloat - certainly if they are flexible with their inventory - but there are still natural limits to such a business, and that makes all the difference in the world.
For most durable goods there's a never ending supply of used goods already; the supply of new goods a couple years later. Once you solve the business problem of building a good funnel for used goods to come into your operation, you can probably boost or create resale value for used goods, indirectly pressuring the original suppliers.