Who's the "we" here? It's already possible to write your own smartphone apps that basically do what supercomputers do. If you work in the field that supercomputers were designed to help (scientists, basically), I'd imagine that they could be very handy, since you could plug your cell phone into a workbench and get results immediately rather than needing to do laborious data entry.

The problem with information is that it's only valuable in context. Scientific computing has given us wonderful breakthroughs, but it needs lab equipment, it needs data collection, it needs trained scientists, it needs theories and hypotheses to test. Computing was developed to break Nazi codes; the Enigma machine may have been the star of the show, but it also depended upon radio intercepts, field operatives, dispatch runners, generals, soldiers in the trenches, people who built fighter planes, etc. - and it obviously wouldn't have been useful had we not been at war. Similarly, most of the consumer data that's vacuumed up is only useful in the context of a task that millions of consumers want to do.

You're right that there's a power imbalance in the distribution of this information. The problem is that this power imbalance doesn't exist because of the information, it exists because of the context in which the information is collected and applied. Find a context where information can let you deliver a better product than competitors, and it's easily worth billions of dollars. But the information on its own - minus the product design, minus everybody else's information, minus the brand name and distribution channel and user adoption - is worth zero. It's already possible to siphon off and redirect all data that is being sent to Facebook & Google's servers on your behalf and save it to use on your own personal behalf. The problem is - it's useless. Without everyone else's data, there's nothing you can do with it.