My desire as a consumer would be to have a system that works across many kinds of hardware from many vendors.
What if you applied that thought to the stage after MC ?
so that things learned by the root could flow to the leaf nodes
I'd like to interpret this as the stuff on the periphery
Tight(er) integration with portable players + tremote and other devices that can interact with MC. Not as much flowing to the leaf nodes but being better integrated with them.
Looking at the poll in the earlier thread, under 40% were interested and 30% were maybe, that could turn out to be 70 for or 60 against. Your odds here seem 50-50 as many ppl like to build thier own.
How do you make a business case with 50-50 ?
Then again its fair to say this is what the more tech-savvy ppl think and the less able could very well think otherwise, your idea to make it easier implies this is the kind of user you want to target.
Do you think you could tranparently handle all the permutations mentioned by jmone ?
..so that it all 'just works'.
The problem is evolution of these parameters in the future and being able to continually cater to them.
One point in favour is you will have a standard configuration build to cater for compared to the much larger mix of machines ppl use MC for currently. You would be pushing out any windows updates or drivers to the users than somebody else, becomes possible to tweak the system for performance and lock it down (virus wise) a little more.
If you pull it of you could have a turn-key solution at a fraction of the cost of a Kaleidscape system