Another area where fuzzy string search is really useful is computational biology. You often have a particular DNA or RNA sequence and you want to figure out what it is. However, the version on file can be different by a few mutations. For this sort of search where the database can be quite large, there are specialized algorithms like BLAST that index by n-grams.

Just thought I'd mention it because if you find it interesting, computational biology is a real growth area right now - the cost of sequencing genomes has been dropping faster than Moore's law, and it's quite possible that within our lifetimes we will be using this sort of technology to genetically engineer people.