You usually see two distinct sales curves for consoles, depending on their popularity at large. The first you see is a curve that starts large, gets larger, and then tapers off eventually (close to a classic bell curve, when viewed on a sales-per-week/month/year scale). The second type starts large and tapers off rapidly (like an upside-down checkmark, again on a sales-per-week/month/year scale).
Historically, it's very rare for two consoles to maintain the first type of curve. You usually see happen what happened last generation: one console gets a bell curve of sales, while its competition gets the inverse checkmark. The degree to which this happens can vary (the N64 and PS1 both got bell curves, but the N64's was much smaller; the Genesis started bell curve and went towards a flatline a year or two after the SNES showed up on scene).
In the cases where more than one console got the bell curve chart of sales for a generation, one always dominated considerably over the other. As well, the future of the curves usually isn't apparent until about two years in; before that, the numbers are too close to each other to say for sure. Then too, there's an important matter to consider: Nintendo's breaking out of the old sales model and seeking consumers who don't traditionally buy consoles. That can and likely will skew things considerably.
While I do see where TheSource is coming from with his predictions, I also think that it's a bit too early to be making such predictions with much hope of accuracy. Past trends are indicative, but not always perfect at telling us what's going to happen. This is especially true when a new marketing tactic comes into play; you can be sure the idea of the PS1 selling over 100 million units would've sounded absurd just a console generation prior when the top-selling system managed just under 50 million units sold worldwide.
Sky Render - Sanity is for the weak.










