DS has had amazing hardware sales so far. Thing is, it'll sell more if it wasn't supply constrained. But, how much more? Is their demand backlog only a few hundred thousand or maybe nearly a million in Japan? What if Nintendo had some way of discovering just how many people want to get themselves a DS who haven't yet because they're not dedicated enough to find one, and they then magically pop some out of nowhere and do a temporary supply spike to catch up with a demand backlog - how many extra will sell?
If that number is large, then that's an even bigger install base earlier on, and a more front-heavy system which means more software sales overall for the full cycle as somebody who purchases the system in the first few years is more likely to buy more games overall than somebody who purchases it in the later years. So if there's a big backlog, that means that there's less software sold as they pick the system up later than they would if supply wasn't a problem.
If that number is small, this disadvantage is minimal and the positive effects of shortages as perceived by casual observers still persist. But just because you've demanded a product longer, it doesn't mean you get it earlier. It's not first demanded, first served; it's first find, first serve. There's potentially a small percentage of those who want a DS who haven't been able to pick one up for several months who then decide to purchase a PSP or something - not all those who want a DS are non-gamers, of course.
But as the DS is such a phenomenon, increasing supply to meet demand could actually be quite a paradox. More units out there means more word-of-mouth and a wider spread of the DS phenomenon, and thus more demand. There's many who haven't yet, but will, Jump In.
More demand means more sales in reality. But in Nintendoworld, more demand doesn't mean more sales, more sales means more demand.
