For libraries using Excel the CSV export often shows some problems: Umlauts/diacritics are broken, because UTF-8 is not detected automatically and the separators are not recognized with all data appearing in one line. What works really will, also with Excel, is the Open Document format that was added for exporting reports. It would be nice if this option could also be added to the items search in the staff interface.
Something needs to be done so that the search results open in Excel well. Parsing semicolon-separated fields is harder than lots of our library staff can handle (not to mention how tedious it is). If Open Document is the solution, I'm 100% behind that.
+1 We often lose a lot of information when doing a CSV export - the data just doesn't export. And as stated in the original comment, the diacritics are broken.