I don't like this idea for many reasons:
* Articles that are updated very often usually got a lot more views than articles that are only published once and never corrected (maybe because they don't need corrections at all);
* Articles with many different downloads may get a lot more downloads simply because some people download all the files. That is, 1 file to download may get 1000 downloads. 5 files to download may get 3500 downloads (I am not considering everyone will get the 5 files).
Also, there is the problem that most popular doesn't mean best article... in fact, many bad articles become very popular and generate a lot of discussions simply because they have problems.
It would be nice if had a committee ( ) of Code Projects MVPs who would give their marks.
Their mark (median of or all their marks or something along those lines) should be combined with votes per category and ratings within given period.