I find it interesting and nice that the 'old 0 star team' wants this run to be published. I personally feel that all of the videos on this site are peer reviewed better than most scientific articles, though the rightful claim to authorship is a bit frustrated. However, the authors of this video do concede that the input (for the movements) is 'unchanged' from the previous video. So, credit is indeed given to the previous authors for their contributions. It's likely incredibly difficult to do, but it would be great if previous authors could be recognized in a more consistent & rigorous fashion for their contributions to the newest submissions. If that could happen, then I believe that a lot of people's qualms about accepting this video, or any qualms about random people picking up any other video to do minor cleanup on, would be eased.
Authorship aside, I agree with a lot of other people that this video is incredibly difficult to watch. Given that the 3 frame improvement is qualitatively non-noticeable, comparing this new video to the old clearly shows that the old video is much easier to follow, understand, and enjoy. Because it was the mission of TAS videos to create something impressively superhuman, time-optimized, AND ENTERTAINING to watch, I believe that this video could easily have its entertainment value improved without compromise in any other area.
If this were a scientific article, it would be sent back for edits before re-submission. Edit the video to fix the camera angles, such that they are as good as the ones of the previous submission or better, and then re-submit it. To follow this ideal, I vote no on this video, and will vote yes when it is fixed.
Good job finding the improvements =).