I don't think this will quite answer your question, but these are things I think about when testing a report to confirm that my results are correct:
How specific are your report specs?
Do you have a test environment that will provide you reproducible results so that you can continually retest reports after changes?
Can you identify a simple population to verify complex calculations? In the report specs, how do you identify specific data elements that are required? Specific calculations?
Identify ALL possible combinations of data that need to be tested.
How do you verify your results ... dump the source data of a small population that covers the different combinations and verify everything manually? Compare to existing reports that you assume are correct?
Just some ideas.... I hope they help.
ttfn, kp
Access to most releases from R52x, on multiple platforms.