What I would hope for is a lightweight (in terms of overhead) OS that works quickly and provide me with the 20/80% feature set mentioned above (20% of the features are needed 80% of the time). With most of the OSes that I use currently I see a lot of overhead and slow response, it seems that as processors speed up the software bloat far surpassed the improved processor power. There will be a market for any device that provides power, speed and is portable with low cost. I would think that Android would be a good contender, but time will tell.
I bet is that as Chrome OS gets more mature, it will start to look more and more like a traditional thick client OS. It'll start with a few things that don't work all that well in cloud or thin-client mode. There will be local backups of data so people can work offline. There will be applications that want more control of the system, bypassing much of the browser component of the OS. Eventually, it really won't be just a browser-based OS anymore.
Back to the story, switch gear from Android to Chrome OS might not be a wise move. It is clear that Android is widely accepted. The gain of market share by Samsung might very well reflect the situation. To both engineering and business, if it works, why change?
To smart devices or PC makers, the question they will likely ask is "What does Chrome OS offer that Andriod doesn't?" The dominated smartphone players are primarily Asian business. They are conservation and will only take risk with careful calculation.
Another question we need to ask ourselves is the availability of apps. If I choose to build Chrome OS, will all the Android apps be available in Chrome OS? If not, as a consumer, I will think twice before switching.
Comparing Chrome OS to Android might be like comparing an Apple to a Orange. To me, they are really built different for different purpose. Google believes they can go to the OS market through the cloud. Microsoft believes they can build the cloud market from the OS. Google and Microsoft have been head on to each other for sometimes. Nonetheless, for Chrome OS to beat Microsoft Windows, I believe it is too soon. On the other hands, the Google domination in the Internet business is difficult to penetrate even though Microsoft bing is doing not too bad.
At the time of launching Windows RT, the challenge to Microsoft is they must do something to earn attention from the market before the tablet market has forgot them. On the day that Windows RT was launched, it drew a lot of attention, good or bad. Nonetheless, imo, it is an impressive device and no doubt nicely built. To me, Windows OS must support wide range of softwares which include any available open source based application and virtualization software. Windows RT is just not there yet. To most tablet users, choices of apps are critical. The more; the better, free or paid. Windows RT is just not there yet. So, what doesn't Microsoft gain by launching Windows RT? To show a commitment to apps developer, to expand their horizon to cloud computing. Who knows! 1 or 2 years from now, Microsoft might indeed earn a good chunk of tablet market which might very well overlap with the PC market.
Android notebooks would be useful for industry if:
*They have a working SAP-client (a real one, not HTML-interface with 19 times nested tables), best would be to make a 2013-web interface with CSS and without tables. Google should push SAP to make it in my opinion
*Remote managing functions, group policies, division between work / private. This totally lacks at the moment. Suited for those pesky admins who need to admin 100+ laptops, should be able to install on Windows and administer Android-laptops from there. Please note, existing "solution" Afaria is easily chosen as worst app all time by users. Good idea to read all the comments (acts as malware / spyware, really instable crashes all the time) and fix it.
*There are cheap rugged versions, current Windows-offerings are ~$2000
*Working WebGL with hardware acceleration would be a pro, especially if there are plugins which handle 3D content
*x86 emulator would be a big pro, even if slow - rewriting a business app will easily cost $35 000 or so
*Virtual Deskopt Infrastructure: Would be a great pro if you can stream a Windows desktop from a Windows server to an Android notebook or tablet, for those companies who need it
*It needs an office suite with the 20% of the functionality of MS Office, especially the 20% of the functions you need 80% of the time
*Should have a good browser, best pre-installed: Stock Android browser is not the best; Firefox and Opera for Android are much better.
I think it was a mistake to introduce the Surface RT tablet first. The Surface Pro tablet, which runs the regular Office suite and can be used in Enterprise networks, would seem to be a good notebook replacment, plus it can also work like a tablet (sans separate keyboard).
Now that the Surface Pro is out, the question is, why consider an ARM-based notebook?