If I read you correctly Lu, what you are saying is that no amount of analysis replaces test results, where the tests were constructed from real operating experience.
BCAR Section A (the "philosophy and approval guidelines") requires a second opinion on all analysis. It is commonly accepted that the second opinion is test results - an approach I find thoroughly wholesome. It isn't mind you a compulsory approach, but not many experienced engineers are truly comfortable with analysis alone, no matter how many people have checked it.
G
"One test result is worth a thousand expert opinions" - Werner Von Braun.