As of 2009, the game industry seems to want two fairly contradictory  things:
Visual aids are nice so here’s one for each of those:
(Ignore the kid yammering over the video, until about 1:10 in, for the quicktime event sequence.) 
We can debate whether encompassing a broader range of human experience is indeed a goal of importance, but if even a God of War game feels the need to have a scene like the one shown above, you might at least concede that it’s something many developers seem interested in furthering.
To cut right to the heart of the conflict I see here, I don’t think we as developers can continue holding our breath and waiting for games that revolve around shooting, driving, running and jumping to someday make a great leap into expressing all kinds of things they were heretofore incapable of.
The problem is that the better versed you are in game conventions, the easier it is to separate the core mechanics of a game from its fiction and theme, and thus say that a game like BioShock is a meditation on free will, the dangers of ideological extremes, and whatever else… despite the fact that you spend about 90% of it shooting people in the face.
On top of simply being good satire, the Onion piece cuts to the heart of that, and reminds us that rest of the world can see this disparity more clearly, ironically by virtue of being less game-literate. For many among the gaming literate, that sort of insight hits pretty close to home.
For a perspective from the other end, I was struck by this comment on io9, a non-gamer blog, from this post about BioShock 2:
I can see how a first-person shooter would be interesting and entertaining, but I would have to fall short of “compelling” when you have to spend that much time, er, shooting.
This person wasn’t being an unreasonable jerk, or advocating the censorship of games. Shooting lots of insane people in a dark, weird place probably just isn’t their idea of a good time.
The common response to this from developers has been things like, “We just need to hire better writers”, “We need better technology”, “We need better artists”, “We need to spend more time planning out our stories”. However, we’ve been doing this for more than 10 years. Whereas if you look at the points where this medium has made the most progress, whenever the expressive capabilities of games have expanded significantly, it’s actually been because new mechanics, or significant developments upon existing ones , have emerged that enable new aesthetics. Those other things are quite important, but we seem to have them covered.
One problem is that, deep down, many designers view game mechanics more as structure (or “form”, if you prefer) than as content, when in fact they are both. If you treat them exclusively as structure when designing, you get all manner of unintended message and context… in a nutshell, ludonarrative dissonance. Which in 2009 means mashing the circle button to overcome an emotional inner conflict. Another designer’s analysis accepts this completely at face value, which if anything demonstrates that this issue transcends our usual valuations of craft and art. It’s almost invisible to us, but quite apparent to outsiders.
So as developers, we need to deal more honestly with the disparity between our reach and our grasp – which is to say, what we tell ourselves our games are about, versus what they are actually about. History will see this decade as the period when games struggled with their destiny in this way.
I’m optimistic though, both because of the progress we’ve made in the first three decades or so of our medium, and because the solutions are right under our noses, deep in the fabric of all games. We must search out, and in some cases rediscover, core mechanics that engender new types of experiences – rediscover, because many have already been done at the fringes, promising yet underexplored. Here are some examples I find especially interesting:
AI Companionship: Holding hands in Ico
You reach out to a non-player character and become connected to them. Suddenly you’re no longer a lone entity; you must account and take responsibility for an Other. Sometimes they’re a hindrance, sometimes a help. Whether or not you buy into the designers’ attempts to make you sympathize, you have a real connection to something that’s reinforced by strong kinesthetics. In Ico, there was plenty of platformy adventuring to go along with this, but it seems inevitable that someday a game will make this its primary emphasis.
Victory via Self-Enrichment: Culture in Civilization
Sometimes you can triumph over an adversary simply by being better than them. Rivals come to view your achievements as an example to be followed. Each accomplishment that enriches you internally affords you expansion and encroachment via indirect force. Tend to your own garden and you will become powerful and influential without firing a shot.
Social Reasoning: Diplomacy
The enemy of my enemy is my friend. Many wargames have a diplomacy component, which gets especially interesting when other humans are in the mix. However in a game where direct force isn’t possible, social standing would be its own capital. This is a large part of why character-driven TV shows are popular; humans enjoy exploring the workings and permutation spaces of social networks.
Hopefully this gives an idea of the breadth of directions available to us as designers. It’s equally fruitful to look to the past, at how certain ideas bubbled up from nowhere to expand the expressive range of games.
Circa 1997, before Thief and Metal Gear Solid, Stealth was one of those underexplored mechanics. Suddenly, as it caught on, there were new play sensations we’d never had before – being some combination of sneaky, clever, afraid, transgressive. It transformed players’ perspectives on familiar game environments. It even brought some new people into the medium.
These are basic changes that everyone feels deeply, from a jaded critic to someone completely new to games. They are interactively “true” in ways that a change in setting can only rarely be, no matter how beautifully realized.
As a medium, we’ve proven we can seek out novel settings, themes, art styles, characters and tropes. We have other media to learn from, after all. New mechanics, however, are uniquely difficult. The only inspiration we can find for them is human experience itself, and then comes the struggle of synthesizing, systematizing and iterating. This is the central challenge of working in this medium, and it’s never been more important that we embrace it.
 While some of this could be explained as the disparity between what game publishers want and what developers want, that might be giving too little credit to the former and too much to the latter. If there were more proven game mechanics and styles that enabled new experiences, publishers would probably sell them. Past a certain point, the burden of proof is on us.
 I want to make it clear that I’m not disparaging GoW:CoO, or speaking in any sense other than constructive criticism. I haven’t played it; in all likelihood it’s a great action game. I’m simply holding it up as an unwitting example of a much more existential crisis in game design today, much as other designers have held up stuff I’ve worked on in a similar light.
 Movement is something that gets re-discovered every so often; Mirror’s Edge being the recent example. Flaws in execution aside, players recognized there was something unique there.