SemiconEng, tooling: Fab 25 is populated with 180nm equipment which used to be used to make processors. Thus their memory manufacturing is already based on last-generation processor generation, so the problems are not as onerous as you assert.
Now, AMD does not use a 'copy exact' since it has a single microprocessor fab, but Intel does not 'copy exact' from processor to memory (so I'm not sure how this point from your post affects anything).
Certainly there are differences, even when the memory processes are based on processor processes. Since migrating processor to memory technology is the fundamental behind AMD manufacturing, I suspect it is a lot quicker (and less problems) than starting from scratch!