It seems to be working on the demo site, we have a file named TRULY_Understanding_ViewState.pdf and I can find it by searching on the title. Note however that it is not going to match on word fragments. If I search for "Understand" it does not find the file because the search index is tokenized on words not on fragments of words. It does seem to recognize the underscore as a word boundary since I can find the file searching for "Understanding" or "Truly". But if the file were named TrulyUnderstandingViewState.pdf I think it would then only be found by the full name because its not going to split on any alphabetic characters when it tokenizes the words.

Be patient when rebuilding the search index, it does take some time for that to complete, it is not instantaneous.

Hope that helps,

Joe