The way I see it, the main advantages of ddt is that you can make testing more user-friendly. In excell-sheets (or just plain tab-delimited textfiles) you write your testcases. And Winrunner interprets the data in the file and carries out the testcases.
(use the ddt_ functions of winrunner to read the files and to make the functions for data driven testing).
e.g. a line in the textfile could be:
enter user <tab> John <tab> Jameson
Winrunner reads in the first column "enter user" and interprets that it has to enter John and Jameson.
Doing all this you separate your testdata from your testscripts, which makes it more maintainable.