Except for the small percentage of people with fluid imaginations, many of whom are borderline schizophrenic, people can't know what they want until they've seen it, or at least heard it described, or better yet tried it out for themselves. The ability to predict what people would want, and be willing to pay for, was no small part of Steve Jobs's genius, and, in the absence of another individual with that same gift, Apple will need a process that can produce results at least nearly as accurate as Steve's intuition did.
I think Apple has all of the elements from which to build such a process already, and only needs to connect them together. Their engineering and design operations already work closely together, each contributing new ideas. To this they only need to add retail; that's right, the stores, hundreds of them, with personnel in constant contact with Apple's customers.
They can't, of course, send product designs out for retail employees to show to customers. Not only would that approach completely negate the secrecy aspect of the company's culture (largely responsible for its mystique), but the feedback it yielded would be almost worthless.
Instead, they need to simply listen, when customers describe features they'd like to see implemented or products they'd like to see built, and pass along what they hear to a group back at Apple HQ, created for that purpose and closely connected to both engineering and design. That group would sift through the suggestions, recombine them, and pass along the most promising of them to product development managers, who might either initiate official projects or authorize skunkwork projects, depending on how close the idea was to describing a marketable product, meaning one that could be built economically enough using available technology to sell briskly at a customary markup.
Even better would be a structure wherein both design and engineering had representatives in the stores, design representatives on the floor and engineering representatives behind the genius bar. These would probably be retail personnel with special training, who would be called over to listen to customers's ideas while other personnel went on with the ordinary business of the store, and these special representatives in the stores could constitute the pool from which the group doing the sifting back at HQ was drawn, providing not only an advancement path (other than management) for retail but a section of the company which, properly led, would gradually become expert at identifying and describing potential products.
It's probably not necessary to have a pair of such representatives at every store, perhaps only 10%. If you think in terms of the adage about saying "no" to 1000 things to find the one thing worth doing, and expect that each such representative will glean an average of one reasonable idea per week, then 100 such representatives should produce one really good idea, worth pursuing, every 10 weeks or so, or about 5 per year. That might seem like a lot of wheelspinning to get a few good ideas, but good ideas are what keeps a company like Apple healthy, and just one blockbuster product would pay for many years of this approach. Moreover, in the meantime, there'd be thousands of customers who left the store feeling as though someone had really listened to them.