Roland's PlugOut is good idea, but of course that has synth electronics in it, and so is far more than a controller. Also they can only control the small number of synths that Roland release that are compatible. Otherwise, today its just case of using automation. NI's NKS standard is pretty slick, but you can only control 8 things. For performance tweaks, handy. For programming, a non starter.
It's easy enough to see the problem of course - every soft synth is different. How can one universal design usefully represent such a wide variety of synths?
THE IDEA
Take a look at this picture, from a beloved broadcast Studer Vista desk.

Look how clear the controls are. Although it looks super-slick, it's actually a pretty low tech idea that - for whatever reason - nobody else that I know of has carried on. It's just touchscreens, with translucent strips and knobs laid over part of it. And that's it. In the UK, I can buy a retail 21 inch touchscreen for under 200 quid - knobs and plastic is pretty trivial. Why not base a universal synth controller on this basic idea?
The way I see it, pretty big would be good - perhaps as big as 21 inch, maybe a little less. A good part of the real estate would be touchscreen - say the lower central part - then the rest would be fixed arrays of knobs and switches. The UI underneath could reflect the UI of the soft synth itself, and of course you could have multiple pages on tags. You couldn't have a literal physical copy of an instrument as the knobs would all be in the wrong place, but you could retain much of the essence. The controller could be designed so that filters are likely to be found in one place, amp envelopes another and so on, but it would all be 100% flexible depending on the synth. Crucially, it would have a good number of physical controls that would - in theory - make programming as good as good hardware.
It would need a new standard to be able to share graphical info alongside all the CC controller - perhaps an ideal would be an extension of NI's NKS standard. (Indeed I think NI would be the best equipped to build such a thing).
WHO WOULD BE INTERESTED?
That's the big question. Is there enough of a market for this? Let's say good size units could be built for $400, and for the purposes of the discussion that it was able to control the main synths you use. Would you be interested?
(just for clarity - this is a purely speculative thread, I'm not interested in trying to design or still less build such a thing. Just trying to get an idea if there would be a real market for it... if there was it wouldn't hurt to drop likes of NI or Arturia a line.)

