Just wanna chime in since people are talking about the subject. Not gonna talk about bias, favoritism, capitalism, etc.
In my view this problem occurs mostly because of the voting system in itself is not robust. For such competition that the entries are judge subjectively, an all-or-nothing system is not something that should be used. It has a lot of flaws, and the end winners might be upsetting for others, as a difference of one vote (especially when there are so few judges) makes a lot of difference.
Not sure how the GMs voted but I just assumed that's how it was done (correct me if I'm wrong).
In my opinion a point based voting system with weighted criterias is more robust and fair to all person involved. There's just no way for emotion, bias, favoritism to interfere when there's a guideline.
To illustrate my point, I've made a spreadsheet (à l’arrache) as can be seen below:
Hope such system (if not already) is implemented for future competitions. Cheers.