re: Time For Plan B At Intel
Intel (or maybe AMD?) needs to partner with Google to build a Nexus 23 - an all-in-one Android 4.x reference computer w/23" touch screen that showcases Android using the x86 processor and graphics capabilities. It would be the most powerful Android device you could buy. The main thing is to tune the board and chips to focus only on what Android could/should do and avoid all the legacy stuff that PCs today are saddled with.
A much stronger case for Linux is also needed, and the non-hobbyist side of the community needs to rally around a single "distribution", at least a core set of stuff that PC makers would not be changing (er, value-adding??) and that device and board manufacturers and software developers could count on to have a stability and a larger market. Part of it means embracing gamer developers (Valve/Steam, EA, maybe even Nintendo) and multimedia (Adobe) so they feel more comfortable and confident in supporting the platform. The overall computing market today is more fragmented than ever, but it's clear that there are some winners, such as Apple and Google, that can be emulated to a degree. Ubuntu/Canonical has some good ideas but they are too small to make a dent, need to partner with chip makers, device manufacturers, app developers, etc. I'm hoping that a brand-new Dell will split off it's consumer division from it's server division and start doing some radical new things. Maybe they need to make the Android 23, or go all-in on Linux. Or both!
ChomeOS is still too limited and can't yet take advantage of the power and breadth of what an x86 PC platform can do. Until it can run Adobe CS and the latest "push-the-limit" 3D games and improve it's offline support, it's not going to go mainstream.
Microsoft needs more competition, and Intel/AMD need more orders. Microsoft will do fine in the end, but they need an even bigger kick in the behind to get moving, and the breakup of Wintel will help do that.