The Wing is not a full open system, is it? How much is the Abacus Pro, upgraded to run the latest motherboards, and OS?

Let's not pretend that Wersi haven't got hugely expensive arrangers on their books... Sure, include the price of its least capable arranger in with the prices of the majors' MOST expensive, but be honest and put the Abacus in there for some perspective!

My Roland can be upgraded with software too... in fact, it has already had two upgrades, for FREE, that added significant new features that weren't on the original at all, plus it has an expansion capability to virtually double the sample ROM (although they have yet to release anything for it, probably due to those reorganization blues). In many ways, modern 'closed' arrangers aren't nearly as 'closed' as they used to be.

Sure, get a Yamaha and your odds of them adding anything at all to the OS are pretty slim, but everyone else is pretty good at adding stuff at no cost.

I'm afraid the open or closed thing, for me, isn't really a cost issue. It's a functionality issue. Until an open arranger BETTERS a closed arranger on ALL fronts (not just a better piano or drums, but everything better, including choice) it is going to face a hard uphill battle. I don't believe that ONE really great piano is enough. There's a REASON there are at least a half dozen different ones in most TOTL arrangers (most MOTL arrangers, if the truth be told, too!). If one was enough, don't you think the majors would have simply used the entire ROM memory for all the different ones, and made one that was significantly better?

I think you are going to see how, once this comes to fruition, that the sound of one drum kit, one piano, one guitar, bass, whatever, after a short time is going to become boring and make everything sound the same. While the exact opposite is what seems to be the (very well researched) direction the majors are going in.

Once again, you are banging into what Dom had problems with. When the styles you play are written for a Yamaha, with 1,500 sounds and more, with 50 kits or more, playing them all into the same generic soundset is going to result in less than favorable comparison. Then add the Ketron styles, the Roland styles, the Korg styles it may be able to play, and you compound the problem. It isn't just a matter of mapping. It is a question of EQ, of velocity curves and filter curves matching, or how each sound interacts with the others to prevent one from overpowering the other.

Then you are going to have to deal with the issue of a style asking for a sound you haven't got around to, yet... What does it do? Do you create custom templates to call up a different sound that you do have, and how well is that going to work?

For the old legacy styles, from back in the good old GS/GM days, yes, you can probably get away with a smaller basic sound selection, because those styles were WRITTEN for an arranger with a small sound selection..! But throw anything from the last ten years at it, styles written for the thousands of sounds in a modern arranger, it is going to be quite a different story.

I still think your best bet is to collaborate with one of the really big sample houses, license a version of something like Colossus, and then spend forever tweaking it so all the sounds are balanced OOTB. Then MAYBE you've got something that a home user can plug his modern BK style, or T5 style, or PA3X style into and it maybe sound almost as good as the original. Without him having to do hours of tweaking to make it work...

Because, if that's what it is going to take to make it sound better than a closed arranger OOTB, he's never going to adopt it.