David Gerard wrote:
I don't see a problem with articles going through a cycle of well-written -> more details -> copyedited -> well-written. Of course, as Ian points out, domain knowledge is important in the copyeditor! (Or at least a talk page note "I've copyedited for flow and structure, please fix any important detail I may have messed up" - an express statement of non-OWNership.)
This would support the notion that we need more sophisticated algorithms for evaluating articles, style and content being only two of the more important qualities to be tracked.
Ec