@ Tom Samplonius
Direct works if all you are selling is hardware, preferably of a specific type. The problem with that is you become a monoculture. And while a monoculture can get big as Dell 1.0 proved, if something changes the environment for the monoculture, you're toast. And something ALWAYS changes the environment for the monoculture.
Once you get into the sort of integrated systems Dell 3.0 is looking to sell, you need channel partners. Your channel partner knows the clients, knows what they need, and establishes the one-on-one not-an-account-number-only relationship you need for the long haul. Yes, it might be more profitable if the sales organization grew that way organically and was all under the same house. It would also be more profitable if I could sell unicorn farts to power your car with no CO2 emissions for a whole year. And those two resource live right next to each other. Back here in Walgreen land you need a channel partner for that kind of integration. Dell will still be able to sell direct to small outfits that don't need that level of integration, or who have somehow landed a genius team to run their IT Support. But the channel were never going to pick them up anyway, so they aren't irritated by that the way they would be if Dell went after their customer list.
What I think you are going to see is a couple of quarters, possibly years, of Dell running losses or at least even more paper-thin margins than they have in the past while they re-org the company. That's the kind of thing you can't be on the exchange and do. You take a 1 to 3 year hit on the profits so you can more quickly get to your full channel goal, then run the next 10 years in a profitability range that makes taking the hit worth it. Gutsy and risky. Wall Street likes the first, but is even more averse to the second than vampires are to garlic and mirrors.