As a general negotiation technique for assigning of responsibility, sure. But management at both Chinese factories is still small and in touch enough to know that if 1M unneeded custom parts are produced, it's a waste no matter how the financial cost is split. American companies meanwhile are so deep in overabstraction and middle management that the individuals are only capable of covering their asses and claim to have solved the problem when the only thing they 'solved' was the ability of the company to perceive the problem. Two years later the supplier goes bankrupt after eating one too many change orders and, despite what the all-important contract says, parts stop getting delivered.
(Not that this is necessarily the Dell case - I'll bet that Dell is on the hook for a certain amount if they suddenly decide to stop using a particular laptop body. It's just that when this philosophy of financial blame becomes your go-to instead of ignoring the financial abstractions and solving the underlying problem, you're begging for someone to exchange your money for false peace of mind)
I don't think it's a false piece of mind. Managers at Dell, for example, know all about the bullwhip effect. They also know that when buying commodity parts, you can generally source from more than one supplier. If you squeeze one too hard and they die, you just get another one because they're lining up to serve you.
You can't "solve" the bullwhip effect - you can lessen it. You can use IT, training, etc. And you can force (through market power) the consequences of it on your suppliers. Doing so doesn't mean that you're blind to the pitfalls of the practice.
None of this has anything to do with being American or Chinese.
The original article is a little inconsistent, focusing on self-oscillation (bullwhip effect), but driving the point home with response time to changing requirements.
Adding bandwidth and damping lessens the bullwhip effect so it isn't the show stopper. It's the response time to changing requirements that's the hard part. If the specifications of a screw need to change and all similar screws are produced in China, having a distributor already stocking (productionUseRate * leadTime) quantities of previously-unneeded types of screws is going to cost a lot more when leadTime includes crossing an ocean. Pretending that the distributor can solve this problem is going to (at best) cost you up front or (worse) end with a large delay as it turns out the distributor doesn't really have that many unallocated screws of the newly needed type. (Actually, connectors are a much better example than screws)
And no, nothing I've said has anything to do with nationality, but everything to do with relative age of economies, their levels of ossification, and fundamentally failing to see through abstractions to underlying realities.
