I came across an interesting passage in a book I was reading (Cracking Creativity, the Secrets of Creative Genius): "A very detailed description of a beach would include every position of every grain of sand. Viewed from a higher vantage point, the details become smeared together, the grains become a smooth expanse of brown. At this level of description, different qualities emerge: the shape of the coastline, the height of the dunes, and so on."
In those three sentences is something that people often ignore in the constant discussion about megapixels and resolution: what problem are you really trying to solve?
If, for example, you really are trying to show your audience every sand grain's position and form, then more megapixels you'll be seeking for a time that is long (stop it, stop channelling Yoda, Thom). In essence, you'll never be satisfied, because you can never get enough pixels to resolve every grain of sand on even a small beach, let alone a big one.
I see a lot of people chasing the individual sand grain. Bird photographers looking for "more definition in the deep in the feathers," portrait photographers chasing "every last strand of hair," sports photographers looking to resolve the logo on the distant golf ball, and so on. If you fall into this category, I have bad news for you. You will be buying new cameras and lenses every time there's even a marginal increase in ability. And don't forget printers and output technologies: you'll need the latest and greatest there, too, otherwise you might lose some sand.
It's usually about this point that I get a Two Defense punch from my readership: "I print large", and/or "I need to crop". The large print defense isn't especially effective because in general the larger the print the further the "normal viewer" is from it. Yes, you can get close and count pixels, just as you could get down on your knees and count sand grains, but is that still a photograph, or are you into some other discipline at that point? I'd argue the latter. The crop defense usually boils down to one of two things: laziness or inability to purchase/carry the right lens. But, it, too, begins to look like the counting sand argument, because when you crop you increase the magnification of the capture area and end up in the pixel peeping category all over again. The reason people that crop want more pixels is to try to avoid the magnification issue. Why not just solve the problem correctly in the first place rather than keep buying new cameras?
What gets missed in all this pixel pursuit is "shape of the coastline" and "height of the dunes." It's the old forest versus trees problem all over again. If all you look at is a tree, you never see the forest. Yet I find that more often than not it's the forest that intrigues me. (If you're into trees, think tree versus branch; if you're into branches, think branch versus wood grain, and so on.)
Let me put it a different way: if you're using a 10mm lens on your D7000, is what you're after the grain of sand or the shape of the coast? That's what I thought your answer would be. When you go after the grain of sand what do you typically do? That's right, you get closer and/or use a longer lens. You don't just turn the Pixel Dial to 11.
"But wait," you say, "there's something about having both the shape and the detail. That's why people use medium format." That sounds reasonable on the face of it, but I'm not 100% convinced of that argument. I've seen a lot of big prints from MF equipment, both film and digital, and in the best ones it's not the pixels that are important, it's the overall increase in acuity. Edges get clear definition. The way our eye-brain system is built, we respond most to changes in luminance (loosely: dynamic range), next in edge definition (contrast), and to a much lesser degree color (hue/saturation). Our brain does not isolate individual rods and cones (loosely: pixels).
Thus, a pixel by itself isn't terribly meaningful, it has to be in support of one of the things that our brain does interpret (luminance, contrast, color). I don't rule out "more pixels," but I also don't tend to over value it. (Note: pros often pursue more pixels to get an advantage over amateurs; they can use any help they can get in these days of ubiquitous imaging.) Be careful that you're actually getting a meaningful benefit if you pursue the more pixel strategy. Foveon users, for instance, will be quick to point out that the acuity is higher for their cameras despite a lower pixel count. If acuity is what you seek, adding pixels isn't necessarily the right solution.
Personally, I try to not look at the individual bits of sand, but try to find the thing I'm really photographing, which is rarely buried in the sand. It always amuses me that some photographers spend enormous sums of money to capture more pixels for portraits, then spend hours removing what was captured from the skin of the subject (without destroying any pixel of hair that might overlap, of course). It's true of landscape photography, too: if you resolve every last little twig and stick and errant candy wrapper, you'll be cloning that last one out. Indeed, more pixels generally means more post processing work involving taking what was resolved that you don't want out.
If you want a real shocker of an experience, try this: take your best image and correctly reduce it to 1920x1080 (~2mp) and throw it up on the 48" LCD or plasma TV in your living room (unfortunately, doing that correctly is an article all in itself). You might find you've already got a camera that can "display big." Sure, you can walk up and put your face up to the display and see the wizard behind the curtain, but at the distance you would normally view that screen from, you're going to be mighty impressed by how it looks. I recently got a chance to do this on a Digital Cinema display (4096x2160, or about 8mp). At the 20/20 vision optimized viewing point, this throws an approximately 55 degree field of view at you. Damned if those 8mp didn't look good, a 55 degrees is huge.
I'll always take more, of course. High tech has this habit of giving us more because the background technologies just keep improving at a constant rate. But I'm no longer pursuing it. Each generation, I pick up a few more details on the grains of sand, but I'm concentrating more on the shape of the shore.