Megapixels? Oh sod off, the megapixel race is well and truly over. Not because the manufacturers have stopped shouting about the number of pixels in their cameras, but simply because photographers stopped caring at some point about half a decade ago.
It's not discussed enough, but one aspect of photography where development is proceeding at an incredible rate, is digital noise. Specifically, the absence of digital noise when you're shooting at high ISOs.
To understand what's happening, we have to take a look at how ISO works and why it creates noise. To simplify it a little: When your camera takes a photo, it combines the amount of light (the aperture), with the duration of an exposure (the shutter speed). It then multiplies the data read by the imaging sensor by a multiplier: the ISO. If you're taking photos at ISO 100, on most cameras that means that there is no multiplier: You get what your camera measures. Swap this with ISO 6400, and you are essentially amplifying the signal from your imaging chip 64 times. Of course; this doesn't just amplify the signal itself, but also any inconsistencies and imperfections in the sensor.
If you've upgraded from an older camera, you can't help but realise how much of a difference it makes. My Fujifilm X100, for example, is a shining beacon of high-ISO performance, and users of high-end SLR cameras like the Nikon D3s are experiencing off-the-charts performance in the high-signal-to-low-noise-at-high-ISO game.
Phrased differently: it's not many years ago that ISO 800 would be practically useless due the vast amount of digital noise, but these days, I'm happy shooting at ISO values well beyond that, and achieve comparable levels of digital noise.
There are two reasons this is happening. The most important reason is that digital imaging sensors are improving their signal-to-noise ratio. How this impacts noise is obvious, really: more precise metering means that when you amplify the signal, you amplify less noise.
The other development is that of better processors in the cameras. The faster image processors (some cameras even include multiple processors in the same camera body, much like you may have a multi-core processor in your laptop or mobile phone) means that the camera can analyse the data it is capturing better. This, in turn, means that cameras can run noise-reduction algorithms, which help reduce the impact of slight mis-readings by the imaging sensor even further.
It just goes to show that a lot of the improvements in photography today are things we cannot measure in megapixels or shutter speeds -- but as photographers, we can take better photos in more circumstances, because the technology is improving in the background. With that in mind, I raise my (very expensive and mind-bogglingly fast) Fluorite glass to Canon, Nikon, and the rest of the gang. Keep up the great work, ladies and gents.