The timing is suspect, as Intel Im sure has its own plans.
My guess is, each card does what it does, or uses its own approach to render, while hydra just calculates which ones do which best, apply them to that usage, and works on down the line, until each card is rendering to its ability.
In a way, this is better, if it is really being done this way, as each company has their own strengths. One card has poor memory management, and slower memory, that card does other things.
What this really does? It wipes out TWIMTBP, which plays into the hands of both ATI and of course Intel, where it lessons both nVidia and ATI to a much lessor extent. A great example is found here:
http://forum.beyond3d.com/showthread.php?p=1320641#post1320641
The discussion is why does the new Batman Arkhum Asylum not have MSAA capable on its new Demo for ATI cards. nVidia totally ignores ATI, working exclusively with the devs at this point, and ATI cards cant use it unless you use a workaround for it. In otherwords, its TWIMTBP at its finest, screw ATI, get in early and tweak the game for nVidia.
And from my link, a typical way ATI does business:
"Originally Posted by Kaotik
And still, giving money and/or help shouldn't mean blocking the other vendor from using features they're capable of
Indeed, it doesn't. One case that I happen to know about was that on securing a co-marketing agreement the first thing the ISV engineers did was take a look at the game code and suggest optimizations that benefitted all DX10 cards, before going on to suggest further enhancements for DX10.1. Would have been very easy to wrap all of those changes purely to the DX10.1 path, but this way was better for the title as a whole."
Now, ATI could have just worked on its own path with the devs, but it also wouldnt have been good for the game overall, for the majority of users out there.
So, getting back to my hypothesis, Intel ruins TWIMTBP approach by working around it, using nVidia when it wants with Hydra, and ATI when it wants for anythting else.
And, like I said, nVidia certainly has more to lose with their somewhat suspect more narrow approach compared to ATIs more broader approach