The best use case I can come up with is that what we're talking about is a cheap scan of the source object, using a scanning technique that destroys that object as it removes each layer to discover the next. The source object would best be made of a very cheap material, of course, and part of its creation might be manual, rather than designed and rendered.
Once the object has been scanned in this destructive way, then it can be mass-produced using the 3-dimensional model now in computer memory--perhaps using much more expensive materials than the original.
The more copies you make, the more value this scanning method might have.
Defining the process as being fundamentally limited to a single copy seems unreasonable.