No, the problem is that some of the export is sorted in a way which led you to think it was a full depth-first playlist-order listing. Which it isn't, was never intended to be, and in the general case cannot possibly be (not while listing each FID only once).

If the output had been more mixed-up right from FID 100, then you would have realised much sooner that this is a "problem" that can't be solved at import time, but only by you (or some other generous person) writing a small program that takes the exported data and generates a depth-first playlist-order listing of it, repeating FIDs when necessary, a bit like this thing. (I'm assuming that the exported CSV does actually contain the playlist structure, i.e. the child-FID lists for each playlist -- I don't have an Emplode install around to check.)

Peter