I can't remember any do/don't do this rules of network about such, so mostly my personal take.
On the outside this seems like simple flip of best of X community wiki concept, that I am fond of. However I think there are major nuances that make negative approach much less viable.
While it's little impact in praising someone for something, criticizing comes with much larger responsibility. There is no value to collection of simplistic rants, such write-ups must be bullet-proof and up-to-date.
What if it's been a year and everything got fixed? Who is going to follow up and clean up answers? What if site gets sued for badmouthing some business (unlikely but not impossible)?
Simply put that would be enumerating badness, which is poor approach to anything. Starting with small selection of good choices is vastly more safe and productive than trying to downsize all possibilities by subtracting known bad picks.
Ten best themes (not happening objectively, for the sake of example) is immensely useful list. Ten worst themes is useless drop in a sea of sub-par products.
Where such content could work is in site's blog. Credible top users taking apart themes to provide excruciating honest feedback is an interesting form and some blogs around do dabble with it.
Only we still don't have a blog. :( A lot of top users seem either highly busy or discouraged with site in recent months. And frankly I'd prefer to see us deal with dangerously growing pile of unanswered questions first.