It clearly states that Proper testing was not done on this problem before the contest started. The worst input for an O(N^2) solution is N = 10^5 and a_i = k\ \forall\ i\isin[1, N], where k is some constant. If they had tested this input against an O(N^2) solution, there would have not been such a huge number of accepted submissions past night.
It is not fair to rejudge these submissions because the contest ended. If they (organisers) had realised about it during the contest and made an announcement saying solutions will be rejudged, it would have been a different story.
I agree that most of the participants have not even attempted by looking at the constraints. Some like me, who at least hoped for a partial score would have attempted with naive solutions and luckily got AC. And as the number of successful submissions increased, others started attempting the problem by submitting naive solutions.
Making the round unrated is kinda not possible because ratings have been updated long back. Reverting all these changes is not gonna help.
This is not against the ones who solved using O(N) or O(N\log{N}) solutions. I highly appreciate their efforts. This is just to advise organisers to properly test the problems and input data before they are put in contest.
This time, the problems in Div 3 were too easy, there were more than 1000 successful submissions for most difficult problem in Div 3., which is kinda not acceptable.
PS: This is just a personal opinion about past contest experience.