Just pulling out some numbers so people don't have to go mucking through the PDF.

LTD (as of Mar. 31, 2008):
DS: 70.6 million, 370 million software
Wii: 24.5 million, 148 million software

Fiscal Year '08:
DS: 30.3 million, 186 million software
Wii: 18.6 million, 120 million software

Fiscal Year '09 (projected):
DS: 28 million (down 8%), 187 million software (no change)
Wii: 25 million (up 34%), 177 million software (up 48%) (Edit: fixed percentages)

Nintendo's Wii software forecast is unreasonably low. They expect to roughly double installed base over the next fiscal year, but they forecast software sales only to increase 20% compared to current LTD. In effect, they're either predicting that new owners will buy much, much less software than current owners, or that current owners will significantly slow their rate of software purchasing. Even if Nintendo sold no Wii consoles, I'd expect software sales to increase, as all 25 million current owners now have a full year to buy games, unlike this past year. Off the top of my head, 250 million is probably a more reasonable estimate.

Also, look for Nintendo to raise their Wii forecasts for hardware and software over the next two quarters at least. Probably more than that, and probably their DS forecasts too.

Edit: Some more analysis.

Worldwide, obviously Wii is already well past Gamecube's lifetime shipments, but it gets interesting when you break it down by region. In the Americas, Wii is still 2.3m behind Gamecube, but it is 1.9m ahead in Japan and a remarkable 3.2m ahead in Others. What is even more remarkable is that Wii software LTD in Japan is only 6.5 million shy of Gamecube software LTD, and that in Europe, Wii is actually already 5.5 million units ahead. This is true despite there being just over half the number of Wii software titles released in each region compared to Gamecube titles. Comparing Wii to Gamecube, the average shipment per title has increased from 100K to 137K in Japan, and more than doubled from 94K to 210K in Others. That's an unbelievable turnaround from last generation.

Even in the Americas, where the Gamecube was a mere failure (as opposed to a miserable failure), software sales have improved significantly. The Gamecube attach ratio looks (10.7) much stronger than the Wii attach ratio (7.50), especially when you consider the fact that there are many more Wii games sitting on store shelves than Gamecube games (and probably fewer Wii consoles than Gamecubes!), thus inflating the Wii number. But the missing factor is that the average Gamecube owner has had the system for much longer than the average Wii owner, giving him much more time to buy games. Comparing the average shipment per title is again illuminating: the average Gamecube game shipped 250K, against 307K for the average Wii game, for a 23% increase. (Please note that I took the liberty of roughly excluding Wii Play. With it included, average shipment is 330K, up 32%.) The really remarkable thing about this figure is that it doesn't come from the top. The big Wii games (Mario, Smash, Zelda, and Kart) have all sold less than their Gamecube counterparts. This means the increase likely comes from the middle and the bottom and strongly implies that the median Wii game has significantly outperformed the median Gamecube game. All this with much less time to sell and fewer console owners looking to buy the games.

Wii is a huge success. Everyone knows that. The real story about it, though, is its unbelievable improvement over the Gamecube in Europe, and its largely unsung software success in the Americas.

Edit 2: Took me longer to write this edit than I intended, and it probably should have been a new post... Oh well.

Just pulling out some numbers so people don't have to go mucking through the PDF.

LTD (as of Mar. 31, 2008):
DS: 70.6 million, 370 million software
Wii: 24.5 million, 148 million software

Fiscal Year '08:
DS: 30.3 million, 186 million software
Wii: 18.6 million, 120 million software

Fiscal Year '09 (projected):
DS: 28 million (down 8%), 187 million software (no change)
Wii: 25 million (up 34%), 177 million software (up 48%) (Edit: fixed percentages)

Nintendo's Wii software forecast is unreasonably low. They expect to roughly double installed base over the next fiscal year, but they forecast software sales only to increase 20% compared to current LTD. In effect, they're either predicting that new owners will buy much, much less software than current owners, or that current owners will significantly slow their rate of software purchasing. Even if Nintendo sold no Wii consoles, I'd expect software sales to increase, as all 25 million current owners now have a full year to buy games, unlike this past year. Off the top of my head, 250 million is probably a more reasonable estimate.

Also, look for Nintendo to raise their Wii forecasts for hardware and software over the next two quarters at least. Probably more than that, and probably their DS forecasts too.

Edit: Some more analysis.

Worldwide, obviously Wii is already well past Gamecube's lifetime shipments, but it gets interesting when you break it down by region. In the Americas, Wii is still 2.3m behind Gamecube, but it is 1.9m ahead in Japan and a remarkable 3.2m ahead in Others. What is even more remarkable is that Wii software LTD in Japan is only 6.5 million shy of Gamecube software LTD, and that in Europe, Wii is actually already 5.5 million units ahead. This is true despite there being just over half the number of Wii software titles released in each region compared to Gamecube titles. Comparing Wii to Gamecube, the average shipment per title has increased from 100K to 137K in Japan, and more than doubled from 94K to 210K in Others. That's an unbelievable turnaround from last generation.

Even in the Americas, where the Gamecube was a mere failure (as opposed to a miserable failure), software sales have improved significantly. The Gamecube attach ratio looks (10.7) much stronger than the Wii attach ratio (7.50), especially when you consider the fact that there are many more Wii games sitting on store shelves than Gamecube games (and probably fewer Wii consoles than Gamecubes!), thus inflating the Wii number. But the missing factor is that the average Gamecube owner has had the system for much longer than the average Wii owner, giving him much more time to buy games. Comparing the average shipment per title is again illuminating: the average Gamecube game shipped 250K, against 307K for the average Wii game, for a 23% increase. (Please note that I took the liberty of roughly excluding Wii Play. With it included, average shipment is 330K, up 32%.) The really remarkable thing about this figure is that it doesn't come from the top. The big Wii games (Mario, Smash, Zelda, and Kart) have all sold less than their Gamecube counterparts. This means the increase likely comes from the middle and the bottom and strongly implies that the median Wii game has significantly outperformed the median Gamecube game. All this with much less time to sell and fewer console owners looking to buy the games.

Wii is a huge success. Everyone knows that. The real story about it, though, is its unbelievable improvement over the Gamecube in Europe, and its largely unsung software success in the Americas.

Edit 2: Took me longer to write this edit than I intended, and it probably should have been a new post... Oh well.