If the code providing correct results does not conform to internal rules or established best practices like DRY, normalizing data etc. I classify it as defect unless a specific reason for that going against principle is in the comment.

>Code is correct if it provides the correct results. The code may not be efficient or easily maintened, but it can still be correct.
>
>>Unfortunately, I have to maintain MY code. I cringe whenever I have to revisit code I wrote a couple of years ago! OTOH, it has worked well as ugly as it is.