Good news: Apple fixed the Tinderbook and sent it right back. Oner week, door to door.
Bad news: The paperwork said they'd put an 80G Toshiba in the machine, which is the right part. But they actually put a 60G Hitachi in the machine, which is the wrong part. It's smaller, and slower. So, gotta do it all again.
Odd news: The replacement disk came with the current iLife suite, but the obsolete MacOS 10.3.4 installed. This means that lots of applications hang on launch. That's not good customer experience, obviously. Worse, the less knowledgeable the customer, the messier and more expensive it'll be to fix.
Management query: Grabbing the wrong part happens, though it makes you worry that the more expensive parts are falling off the back of a truck down there in Memphis. But I'm having trouble imagining why you'd load up replacement disks with mutually incompatible software versions. Slapping new drives into machines has got to be one of the most common things they do at the repair depot, and I would expect that software loadup would be essentially automatic.
This is an easy system to engineer right, but the fault can be hard to detect from the outside: the service guy is in Memphis, the call center is Somewhere Else, and the support costs for the bad install get charged against one budget while the premature disk failure is charged to another. One bad tech, or a good tech who was handed a bad disk, could cause thousands of dollars of support calls. It'd be easy to see the problem in advance, but could be a real headache to sort out if you're just a manager wondering why call volumes are up a tick.