The "base" in the context I'm meaning are those people who are generally inclined to buy one console over the others. GCs 20M represent the Nintendo core: being those who will always buy Nintendo in good times and bad. It's possibly to lose a lot of the base if you screw up badly. It's not really possible to lose the core. Nintendo screwed up badly with GC so they pretty much lost all of their base. They got it back with Wii.

Sony established a large base because they sold a butt load of PS1s and PS2's. They sold more PS1's than any non-handheld Nintendo console, and then they went one better with PS2. By comparison Nintendo's best selling non-HH console before Wii was NEs at 61M. PS3 has sold more than NES by 5M units. And it looks like Wii won't beat out PS1. So PS3 has sold better than every non-HH Nintendo other than Wii, and Wii has sold worse than every non-HH PS console other than PS3. What does that tell us? The base for brand Playstation built up pretty rapidly over the course of 2 generations. The base for non-HH Nintendo cannot be said to be higher than Playstation on the basis of Wii's performance. It might be higher than PS if the first time ever console buyers who bought Wii can be consisdered part of Wii's new base. But that's far from certain. Why would a person who bought Wii for Wii Fit feel compelled to buy a Wii U? They have no particular brand loyalty to Nintendo or Wii. The evolution of fitness software on consoles is arguably on Kinect, so the Wii Fit demographic may have already gone to 360/Kinect. Or they may be more inclined to get 720/Kinect 2.0.

Estimating the base for each console brand is pretty much a guess on my part. But estimating the base will be connected in some way with the sales history of the brand. We have a 3 generation history for PS brand and it averages out to about 100 million per generation. But you've also got to look at the brand's low point when considering the base too, and the context of that low point. PS3 had a bad generation, at least until 2009. It was total doom and gloom with predictions of 10-15M lifetime sales. But in spite of that PS3 has managed to have a reasonable generation in terms of units sold. That suggests the brand's base is pretty close to PS3 sales. Not all of PS3's sales will be to its base, but most of them will be. 60M is optimistic for me perhaps, but 50M is very realistic for PS4 even if all Sony does is slightly change the controller to improve some ergonomics and of course produce a PS4 with the rumoured specs or thereabouts.

The Xbox base is the least discernable as it's got only 2 generations of information and they've established no pattern, and their first generation was rather lackluster, unlike PS. It's base might be 25M or it might be 50M. But at this point, because Xbox is so strong in USA (i.e. it has a solid base there now I think) and UK I think 50-60M is well and truly on for 720. So maybe I should express things in a range, like 50-60M for the 3 consoles and hence 150-180M consoles are pretty much pre-sold.

You gotta remember that most gamers are "nomads", they go where most 3rd party games is, as has been apparent through the whole og VG history. NES had massive 3rd party support. Snes had even bigger 3rd party support. PS1 snagged all the 3rd party support, and keept it with PS2. Then along comes Xbox360; getting equelly as much 3rd party support as PS3, and it's pretty much a draw between them now.

That's what might present itself as a problem down the line for WiiU - if PS4/720 are roughly the same power, and both based on x86 CPUs and ATIs GPUs it will be breeze for 3rd party developers to make multiplats in next generation for those two system (unlike what we had at the beginning of this one, due to PS3 architecture). Now, I have no doubt that developers using current gen tools and engines will have no problem making solid and realativly easy WiiU versions once 720/PS4 hit the market, what I fear is what will happpen when they start using next-gen engines and push them to the limits - how hard it will be for UE4's "real-time everything" pipeline to downscale or bake to UE3 for example. I'm certain that it probably can be done for lot of future titles, but if it turns out that it requires significant time/money, desicion will probably be made depending on WiiU's user base at the moment - and that user base must be significant if they want for major 3rd parties to continue supporting them by investing additional resources in WiiU ports. That said, I don't see much people jumping to WiiU and leaving their PS360s behind (WiiU is not really showing generation leap against PS360), once 720/PS4 come that's where they gonna migrate to.