For the most part we aren't talking about computational biologists but experimentalists using Excel. People at the bench need to collect their data somehow, and using Excel for tabular data and Word for text data is just what they know. Typically they then pass these files over to computational biologists for analysis. Yes, it would be nice if they would use more appropriate tools, but I know from experience that the typical result of trying to teach them better tools is the experimentalists just rolling their eyes and saying that they don't have time to learn some nerdy program because they have experiments to run.