The exact value of this absolute zero, relative to Celsius temperature has been measured to within a few microkelvins (it is actually defined as 0K, which is defined to be -273.15 C) It was well established IIRC in the late 19th century by extrapolation. It actually cannot be reached (this is one of several alternative versions of what is called The Third Law of Thermodynamics: "By no finite series of processes is the absolute zero of temperature achievable.") In its strongest formulation, the absolute zero of temperature isn't really defined in terms of temperature, it's defined as the temperature at which the entropy of a perfect crystal is zero.
A much better discussion of what is going on here than is stated in the article is in http://en.wikipedia.org/wiki/Absolute_zero in the section under "negative temperature." It is brief and accessible to the layman.
The beauty of the macroscopic state variables of thermodynamics is that they don't depend on any underlying theory of matter: none of the theorems or results of classical thermodynamics were changed when classical physics was modified by relativity, and none of them were changed with the advent of quantum mechanics. If quantum mechanics was overthrown tomorrow, macroscopic thermodynamics would still be entirely valid and not a single definition, result, or equation would change.
What's frightening is how well they agree over how wide a range of conditions.
The odds of that must be around 1720, due to the 2nd Law of Thermal Documents /crevo-thread>
Cheers!
You have done a very good job of making this rather mis-named concept of “negative temperature” accessible to people here. Thanks.