There was merit in the concept of file deduplication, especially in distributed file systems, as I'm unsure if this existed as prior art since I can't find any reasonable literature before 2000. I'm pretty sure someone had created a file deduplication script in UNIX at least 20-30 years prior, but I can't find references to that work, so I'll assume none exists and give this patent the benefit of the doubt. Though the 'uniq' command in UNIX could be trivially used to implement this if you consider the use of MD for file uniqueness checks, so even that credibility is questionable. For example, fill a file with contents such as:

<md hash> <filename>\n

For all files in the filesystem, then uniq on the first N characters where N is the length of the md hash, and use -d to display duplicates, then remove those files. This can also be used as the mechanism for "attaching a substantially unique id to a file" as well as half of the claims in the first patent, which to me makes a strong argument that it was obvious.

However, everything else about it describes an obvious application and implementation of a hash table using a "substantially unique" hashing method, which could easily be described by MD5 which was published in 1992. For that reason, at least half of the claims in this patent are covered by prior art, and so the patent is, IMO, far too broad to be considered valid.