The main thing here, of course, is the sudden focus on graphics. It's a novelty that, once the graphics have gotten so good they can't improve much anymore (a boiling point that I think will be reached within a couple of years), will wear out and attention will return to gameplay and story. This is a growing period for the industry. It's having teething pains, that's all.
I agree with everything you say here, but I don't think graphics will reach their limit that soon. Well let me rephrase that. I don't think
visual realism will reach its limit for a very long time. We've still got to overcome the
Uncanny Valley. Personally, I think that's going to leave a lot of room for improvement.
So let me explain what I mean by visual realism vs. graphics. A graphic, by its definition, is a flat, motionless picture. In computer games, we draw those pictures on the screen at 30 times per second (more or less). This creates the illusion of motion, and the overall experience we have constitutes the visual realism of that game. Graphics may be nearing their limit, yes, but the realism of those graphics in motion has a rough road ahead.
No matter how good we make a computer model of a person, even in film special effects, we are able to discern it as a fake once it is animated. In the pretty far future, only when it is possible to create and model characters based on musculature, bone structure, etc, with their movements entirely AI-controlled (as opposed to motion-captured animations), we will cause them to act in a lifelike way, and will truly overcome the valley. We will actually think and feel that the character on the screen is a living, breathing person. What a trip that will be.
Just think about the complexity that goes into the motion of a human... and how do we duplicate it to trick the viewer into thinking that person is real? They must breathe properly, speeding up when exerting themselves, slowing down when resting. The intake of the air swells their lungs, which react mechanically with their flesh to produce visible motion to an observer. When they talk, move, or just stand around, they are idly (and subconsciously) shifting their body weight, moving their appendages, and focusing their eyes on different things.
Several developers have attemped to emulate these idle motions as preset animations, and have in fact animated entire characters based on motion capture alone. But think of the possiblities when motion is controlled entirely by AI!
AI says "move your hand upwards." The virtual bicep flexes, along with the anterior deltoid, and the rigid bone structure causes the arm and hand to move in the desired way. But during this motion, a projectile comes in at the character, and the AI redirects the hand to shield its eyes. All this time, the rest of the body is acting in a realistic way.
Maybe I'm going a little overboard, but what I'm trying to get at is that even when we reach the point where a stationary human model looks real on a computer screen, once set into motion, we become aware that it is in fact a fake.
But someday, we'll have the processing power (and programming capabilities) to allow AI to bridge the gap.
I, for one, am excited.
But, that's the long term. As for the short term, I totally agree with you. I sense, and hope for, a gameplay revolution in the next decade. A renaissance, if you will, of the golden age of gaming. I have a dream that someday, there will be no more "shiny," "purdy" eye candy to go around, and we will be forced to judge games not by the color of their skin, but by the content of their character.
...in the meantime I'll be playing super nintendo.