TTH is an excellent feature if you want to make sure that you aren't downloading a fragmented file from a user but the correct file.
File corruption happens alot is when users download a file from multiple sources and one source leaves then he resumes from another. This often results in the content of the file being misplaced and different although the file size is the same as the original.
I always try to pick the files which contains the most similar TTH lines, which means if everyone has it its not corrupt
. But then again from time to time this doesn't hold true. I've seen multiple users with the same TTH and the files were corrupted. Maybe they downloaded something big and found out its corrupt so their too lazy to redownload it or delete it 
. This can be a pain if your downloading something over 500mb and have to redownload it after you find out its corrupt.
One method to further prevent this from happening is to implement a CRC checker. Although its NOT GUARANTEED that the file your downloading isn't going to be corrupt because in one case, not all authors will put the CRC code of the file your downloading on their website(althought you can ask).
The author or user of the original files sometimes will put up a CRC code on their website to ensure that when you download the file, it's the same as the original and not corrupt or different.
TTH + CRC = almost impossible to download a corrupted file
.
Just a suggestion.
-stealth