Often when reading reviews, I tend to see that a lot of the time, it is the subjective experiences of the reviewers that determine the score of a game.  Some try to convince you that even though a game has some technical flaws, it makes up for it with something compelling either in narrative, gameplay, or in another area.  
 
I don't think they have the authority to rule out any technical flaw based on a subjective experience.
 
Most people seem to be under the impression that games in general are just as subjective as movies, music, books etc.  This is a false notion.
Most people think that a game is still good if it provides subjective experiences that will be good enough to ignore the bugs.  This is simplifying the medium too far.
 
A game is made up of two very distinct parts - The technical half, and the artistic half (for lack of a better name sorry).  All programs that perform functions are judged on how well they perform the tasks they claim to do, as well as the quality of the overall package.  For the most part, programs are easy to review without bias, because they either work well, or they don't.  Games, believe it or not, inherit this trait to some extent.
 
If a game works well, then the final review will be a subjective one, as the only thing that can make or break the interest of consumers is the feature set, narrative (if any), setting and artistic direction of said game.  No one can definitively call the game poorly-developed, though can fairly say that they are not happy with the feature set options, or subjective elements like narrative.  Even in the most extreme of situations where games seem to have unquestionably horrible characters, art direction, narrrative and/or setting, those variables are still within the realm of subjectivity, and isn't quantifiable by any possible means.
 
On the other side though, if a game was poorly made and is riddled with bugs or experience serious performance issues, we can objectively say that a game is poorly-made.  If there are problems with gameplay, there is no way to ignore it or see it in a nicer light.  Same goes for frame rate drops, incorrect values, poorly-calibrated control schemes, visual artefacts that were clearly not intentional, bugs that will hinder or even halt a player's progression, corruption of data, failure to retain a recent state of a player's current game, sound issues and many more that I can't think about right now.  An accumulation of these problems can greatly determine the quality of a game on this aspect of a game, and often enough it is poor quality control that will gaurantee a negative impact on a game. 
 
Most people may argue that bad forms of the subjective elements of a game will also guarantee a games demise.  I would then have to remind these people that no subjective element can ever "guarantee" anything.  They can be used to make a calculated guess, but guesses are hardly definitive in the first place, let alone these elements.  They were never able to guarantee, and they never will.  So many people rate Starcraft 2 so highly for being an excellent game, but in the end the only thing that can be gauranteed is that Starcraft 2 is a well-developed game at its core.  You cannot convince people who were never interested in RTS games to like Starcraft 2, simply because it isn't their cup of tea.  However they cannot deny that the game is polished and well-developed to the point that only subjective elements will determine whether or not it is worth their money.  I could say that I do not need chat rooms, but that is my subjective opinion that will differ to another person who might want the feature added.  The addition or removal of a feature does not affect the games overall performance and reliability to perform as best as possible to make the subjective experience an error-free one.
 
As games continue to evolve, there will be many more subjective experiences that will be made possible by the advances of technology, as well as the continuous push for innovation by creative minds.  These will result in many more discussions revolving about many topics, like artistic quality, questionable feature additions, removals or alterations, the dollar-per-hour debate and whether or not it is an acceptable line of thought for deciding the worth of a game, and many others.  As we explore these questions though, we must never forget.
 
Games are still programs, and they cannot be completely subjective until we can guarantee a level of quality in the technical aspects that will prevent any game issues from occurring in the future.  Theoretically, this is an impossible variable to keep constant, and thus there will always be a level of objectivity that will definitively separate good games from bad games to some reasonable extent.