I haven't watched this run, nor have I watched the published run, but I thought I'd throw in my two cents for this type of condition.
There reaches a certain length of movie where the argument moves from "faster is better" to "entertaining is better". I don't know where this line is drawn exactly, but a run like Donkey Kong (NES) derives most of it's entertainment from the sheer speed it's beaten, and not much could be done to make it more entertaining while still beating the game in a reasonable time.
However, the longer the movie gets, the more chances occur for small entertaining bits to be thrown in. Because of the length of movie, these little bits don't affect the overall enjoyment that comes from speed, since a few seconds missed over an hour long movie seem miniscule. This can backfire though, when a long movie misses chances to throw in these little entertaining bits, or worse, makes blatant mistakes, which can certainly cause the movie to become less enjoyable. And this movie seems to have many instances of that occurring, from what I've read.
So, if you watched this movie, and thought it was enjoyable because it broke a record and was fast, then that's your opinion and certainly a valid one. But if you watched it and the lack of optimization caused you to not enjoy it, that's your opinion and also a valid one (although it is rarer for a submission to pop up with these types of problems).
For a longer look at another movie that had very similar arguments, see Vatchern's RCR submission
http://tasvideos.org/forum/viewtopic.php?t=5086
So, if enough people found that the lack of optimization made the movie not enjoyable, it shouldn't be published. Any arguments about whether the original should have been published are moot, because it's been done and is not going to be undone. See the 3 NES Track and Field movies for proof of this. So yeah, my two cents about the issue in general.