Horatiu argues that this regime may be described by the effective pion model - and the black hole is a nonlinear soliton of this pion field. From the pion model, he can calculate the temperature of the soliton, and the result is
- 176 MeV times "a"
So this is what the string theory dual of QCD predicts for the temperature of such black holes. Can we test string theory experimentally, at least this string theory dual to QCD? Can we produce these dual black holes? Yes, this is what RHIC is good for. So Horatiu looks at the RHIC data and he sees that they have produced fireballs whose temperature is
- 176 MeV
QCD, string theory, and its uniqueness
Let me clarify the situation a little bit: Horatiu's ideas and similar approaches, even if they work, are not testing the "full" string theory. They are testing some particular mathematical methods applied to approximate descriptions of Nature, such as QCD. The dual dynamics is a not-quite-geometrical theory in the anti de Sitter space, although many of the conclusions we would make in a geometrical anti de Sitter setup will be qualitatively correct.
QCD is not the full theory of Nature, and correspondingly its dual string theory is not the same type of a solution of equations of string theory that we consider to be the theory of everything. The realistic backgrounds of the full string theory can predict everything, including gravity and black holes in the actual spacetime - which are different objects than the "dual black holes" in the "dual string theory to QCD".
For example, using string theory I can predict that Peter Woit would only have read the beginning of this article, before he will submit the first comment under this article. Moreover, in his comment, he will state an incorrect conjecture that I think that string theory is just a "dual of QCD". I definitely don't think that.
Note that it is easy to make physical predictions about simple physical systems - for example low-energy gravitons in 11 dimensions, fireballs, or Peter Woit - but it is harder to calculate the properties of more complicated systems, such as the proton.
I also predict that in his second posting, Peter Woit will ask whether the existence of the "string theory dual of QCD" contradicts the uniqueness of string theory. I actually think that this is a good question. But it seems that the answer is "no": these "different string theories" are just different vacua of the same theory, and it is conceivable that some of them are not solutions of string theory at all.
The dual string theory to QCD is moreover a very peculiar type of background - as string theory, it is a highly curved five-dimensional space. We would certainly not consider this background to be a full description of our Universe including its gravity simply because the characters of the gravitational forces in these two backgrounds are very different. But it is not quite a trivial question whether the dual string theory to QCD is a solution to the same theory that also describes the full world around us. Whatever the answer is, it can't invalidate the uniqueness of string theory: on the other hand, it is - of course - an example of the multiplicity of solutions of string theory.