Keep miniaturization in perspective. "Smaller is better" doesn't necessarily apply.
When it comes to changing the face of power management, I can see why there's quite a buzz over the slew of new microgenerators and energy harvesting-and-converter sensors now in development—some of them regarded as taking the first steps toward a perpetual motion machine. But I'm just as confused by the mixed messages I get as to the reasons for the excitement. One is tied in with their physical size. Indeed, I'd expect a device that puts out nothing to be extremely small! It makes me think about miniaturization in power sources and components in general and the bedrock fundamental that "smaller is better."
But maybe it isn't. As an aside, the new micropower devices, drawing on energy from the sun to RF to motive (pedestrian) sources, will be useful in extremely low-power remote systems and where you can't access/replace traditional batteries. And I'll agree we've gone maybe as far as we can go in improving the volumetric efficiency of batteries. But if any of the new sensors are intended to replace batteries (some are), then it might still be better to go with traditional sources when you can to get more power in less space. After all, if the sensor is going to take up the side of a wall or is going to be buried in it, or is part of a pedometer, we're not talking about power management as such but devices that are suited to specialized applications.
But we digress. Physical size is a part of any product's advantage, you say. Yes, the maxim "smaller is better" has held sway for a long time. But the clock runs down on bedrock beliefs all the time, even when it comes to the natural laws of the universe here in 2007. And when it comes to power components in general, smaller ceases to be "better" when it introduces complications, such as heat dissipation in constrained environments. That's caused more trouble, from exploding laptop screens to power supplies, than any other. Is it worth
Or take power supply capacitors. Their volumetric efficiency, too has greatly improved. But when they shrink to beyond what you need, they often won't handle the same surge currents. That's why you often have to add a current-limiting board to work with them. It's adding complication for nothing; better that the caps were larger to begin with. Apart from power, there are apparent exceptions to the rule. Like software. Smaller programs are often more reliable than larger ones because they introduce fewer complications.
Does a smaller device or system give you the same or greater performance and more flexibility? And do you really need the increased flexibility? If not, take a second look. Go ahead, make it a little bigger. In many cases, you'll benefit. Because there's a law that's been valid longer than any other: simpler is better.