A real world application of cross join is to generate the results of applying databased rules to the row occurrences they apply to. For example, I have an application that tracks the set up of standard AD Groups for database schemas. I have the rules for these groups modeled in tables, and then cross join the rules with the schemas to generate a result set that applies each rule to each schema. I output this result set to a tracking table so I can track the set up of all these AD groups. This of course is only effective because there are a small number (less than 100) of schemas for which to manage a small number of AD groups (around 8 per schema).