One of the big problems with digital music is the absence of tactile interfaces. Once you add a decent-sized screen with multitouch and perhaps pressure sensitivity, a lot of limitations disappear, and you get a digital instrument you can play, rather than just poke a cursor at.
I'd expect similar changes in other areas.
But a lot of computer-think has been driven by business - spreadsheet, power point, document server, yada yada - and businesses aren't fond of change.
There are two issues - physical interfaces, and mental metaphors - with some synergy between them.
I'm sure better metaphors are possible, but they'll be much more acceptable if you don't have to drive them with a mouse and keyboard.
The digital side of semiconductors is just about tapped out in terms of transistor size and speed, and hardly anyone has needed more than the most basic processor on the market for their needs for the last 10 years. On the software side I think Adobe's model is the trend: break backwards compatibility and force consumers to buy new versions of software that don't do anything new (at best).
you are the media you consume.
or this
with a completely configurable panel, which is potentially context sensitive, or at least switchable, and can be laid out ad lib.
The new paradigm comes from seeing hardware as a tactile interactive object that happens to do some computing, rather than as a computer that happens to have a touch screen.
The iPhone already does some of this, but it's too small to do more than play games on. If you scale up the size, the possibilities get more interesting.
I was working as the Systems Architect for a system not too unlike what you're describing back in '89. Ran into an insurmountable problem and had to round-file the effort. Can't remember what the problem back then was.