It seems like publicizing some preliminary results of the study before the study is complete would put the legitimacy of the final results very much so at risk.
After 80k votes, it seemed prudent to give those asking "so what?" a response.
If (as I'm sure you do) you have the times along with the votes it would be cool to see if there was actually a bias when comparing the before response and after response datasets.
Yes, but so would (presumably extreme) sample bias.