I have a lot of issues about some ideas on here, particularly multi-task devices taking over all other devices and cloud gaming. I wrote a long post on both, but I'm going to focus on cloud gaming.
The assumptions which make cloud gaming feasible are poorly made. The biggest wrong assumption is that the needed data bandwidth will be available in the near future.
Assumption: increases in bandwidth and data use will stay the same. Anyone who thinks that there is infinite infrastructure for bandwidth clearly has not paid attention to the ATT iPhone problems and the throttling issues with ISPs. Wireless providers are trying desperately to upgrade to 4G LTE networks, but that upgrade is expensive and time consuming. True 4G standards are 10 times faster than what wireless providers call 4G. ISPs are having no better luck making their networks faster. Once we reach certain speeds, data network use starts interfering with control network data. I'm not going to explain all this here, but look up these issues on your own time. Bandwidth expansion over ground and wireless networks cannot grow as fast in the future as it did in the past decade.
Why is this important for cloud gaming? Have you played cloud games? They are laggy. Most games are almost unplayable when there is a 1/4 second or longer delay between control and action. Without higher bandwidth, HD games will reflex controls cannot be done through cloud gaming.
Another problem with cloud gaming is the cost. I won't try to guess what it will cost the consumer, but lets look at what it will cost the various providers. Console/PC gaming requires the consumer to buy the console/pc, while cloud gaming providers must buy the gaming devices themselves, house them, power them, maintain them etc. That cost is forwarded to the consumer the same way that the electric costs of gaming are the responsibility of the consumer. However, the bandwidth used by the provider is enormous and that is a new cost that does not exist in traditional, home-based gaming. That cost too is forwarded to the consumer. At the same time, the consumer must pay for the receiving end of the bandwidth. Essentially, the consumer has to pay for the transmission of the data, and the reception of the data, so the consumer pays twice the interest cost.
Remember that cloud gaming companies do not pay for anything on their own, it always comes out of the pocket of the consumer in the end. Subscription costs for a profitable cloud gaming will be astronomical unless they can make the hardware costs less by industrializing it. Maybe big, multiple-game systems will run more efficiently than single-game systems. But who develops those machines? Where do those development costs come from? And maybe multiple-game systems cost more to run after all.
Cloud gaming is a neat idea, but in actuality it has a long way to go. I'm not going to invest in it, but then again neither are any of you. Console gaming is NOT going away, get used to it. I'm on the fence about PC gaming, but it's pretty expensive compared to console gaming... so the only reasons I can see for PC gaming staying alive are 1) better controls and 2) people who see console gaming as too mainstream/not hardcore enough (IE, people who see PC gaming as a cultural/personal statement rather than through a cost-benefit analysis).