I think the biggest issue is not the objective winning criteria, rather I think it's the narrow definition of this site as a whole. Perhaps the the scope of this site should be expanded, not to replace the rules, but to include more than golf tips, code golf, and programming puzzles.
By the definition of its own name, "Programming Puzzles & Code Golf" does not include challenges where answers are picked collectively by the users themselves or by some subjective winning criteria (judged competitions).
Yesterday, I started the question Hello World in 1024 characters. It was my first day perusing this SE site (and I apologize for not being more familiar with the rules). My first impression was that this site was for having fun with code, first and foremost, and that was what drew me in. My second impression was indeed the lack of participation in many of the questions, as well as the low frequency of new questions being added. In an attempt to stir up some activity, I started a question that I thought was simple enough to tackle in a reasonable amount of time, would include a plethora of languages for diversity, and encourage active gamesmanship while people tried to out-wit each other with creativity.
The problem I see with code-golfing is that not all languages have a chance at winning, leaving a distinct advantage to individuals with particular experience. In addition to this, within whatever language it may be, there is ultimately only one true possible shortest code, and the first person to post it wins. This leaves very little incentive to a vast number of coders to toss their hat into the ring, so to speak.
I do not think there is any doubt for wanting to increase the activity here. I'm fairly certain we would all love to see more new faces, and more of the old ones. By my thinking, the best way to accomplish this is by having questions that inspire the community to post answers, any one of which having the potential to be selected as the winner, instead of intimidating those lesser experienced of us from posting anything at all.
Looking at the answers to my question, it is clear to see that the range of skill is immense, from the spare time hobbyist, to the code-for-a-living professional. Had I posted a challenge that I thought would challenge everybody, I would have left a large portion of skillsets on the doorstep. Conversely, if I had created a moderate challenge that would be geared towards those further towards the "hobbyist" side of the skill-scale, the big dogs would blow it out of the water, which would ALSO discourage lesser experienced coders.
The inherent problems that arise from voted or judged questions are based on the subjectivity itself.
- The OP may not be familiar with many of the languages he or she is
likely to come across, and therefor picks an answer written in their
language of choice.
- The users who are doing the up-voting may not be
up-voting based on the intended criteria set in place by the OP.
- The OP may never even select an answer, and without an objective winning
criteria, nobody ever finds out who won.
If we can address these three points, I think we'll be onto something.