But if you buy finished iPads from Foxcon then they don't count until the finished iPad is shipped out the door. Whereas Dell might buy in NVidia cards, Intel CPUS, Corsair memory and assemble it itself so that does count.

Although Apple are good it's largely a trick of how they manufacture. If Dell had it's assembly plant as a subsidiary company and wasn't billed for the computer until Fedex shipped it - they could have a 1minute turnover.

Sure, but don't try to hack this metric too much. Manufacturers use it and don't try to game it. They actually want to improve it. Apple partners with Foxcon to get this number at 5, and wouldn't have it at 5 if they didn't believe Foxcon could manage demand. Apple would maintain inventories in their warehouse if that was the case.

This metric speaks as much about Foxcon's success as it does Apple's.

Yes this is mostly down to Foxcon and that Apple sells a relatively small range of machines with little customization. And it only counts for Apple's own stores, not machines on the shelf at Staples etc

And if you take this metric to extremes then something like the Morgan car company - where there is a multi-year waiting list for their hand built sportscars - is the most efficient company in the world !