Though I am not familiar with this tool, one option would be to run the models to completion and use the Results Comparison tool. Though it will not tell you differences such as which rain event used and such, it should give you some indication of at least where to start looking. Otherwise, it might just be easier to pull all of your dbf information into Microsoft Excel, or any other spreadsheet program you may have, and use arrays or pivot tables to analyze the differences. However, as you said, it would be a table by table comparison. Fortunately, there would only be able 10 - 15 tables you would need to look at depending on what the expected differences would be.