What is not mentioned about the Nvidia GF100 - power consumption, tesselation and manufacturing

The hype and FUD machine is ratcheting up to full speed for the released-sometime-this-spring Nvidia DirectX 11 card, the GF100, a.k.a. Fermi.  I’ve been reading the recently reported Nvidia specs and then someone pointed me to this article on SemiAccurate:  Nvidia GF100 pulls 280W and is unmanufacturable.

The gist of the article is:  The GF100 was never meant to be a graphics chip but rather a GPGPU compute “generalist” chip. In this new role as a GPU, the GF100 too costly to manufacture, too power hungry, and really is not equipped to do real-life gaming or professional rendering (but not bad in specialized benchmark comparisons).

Read the articles for the specifics, but for me (being a fan of tessellation), the following stuck out: “On ATI 5xxx cards, tessellation performance takes almost no shader time other than dealing with the output just like any other triangle. On GF100, there is no real dedicated hardware, so the more you crank up the tessellation, the more shaders you co-opt.”.

I am guessing the level of FUD will only increase.  It will be interesting to see where the market goes.