However, a new theoretical research has taken a different angle. Two physicists in Austria has published a paper showing that the emergence of classical observation can be also be obtained without having any decoherence effect, but rather due to the "coarse-grained" measurement that we make. A review of this work was reported in Nature Daily News (the link may be available for a limited time and may require registration and/or subscription).
Johannes Kofler and Časlav Brukner of the University of Vienna and the Institute of Quantum Optics and Quantum Information, also in Vienna, say that the emergence of the 'classical' laws of physics, deduced by the likes of Galileo and Newton, from quantum rules happens not as objects get bigger, but because of the ways we measure these objects. If we could make every measurement with as much precision as we liked, there would be no classical world at all, they say.
We know that "size" isn't the issue here, especially with the recent SQUID experiments of Delft and Stony Brook. However, the conventional thinking is that the larger the size, the more difficult it is to maintain coherence of all the parts of the system. What the new approach here has tried to explain is that with the larger size, the precision of our measurement also tends to get worse. Unfortunately, their proposal to measure and detect the quantum effects on large system appears to be rather daunting, if not almost-impossible.
Kofler says that we should be able to see this transition between classical and quantum behaviour. The transition would be curious: classical behaviour would be punctuated by occasional quantum jumps, so that, say, the compass needle would mostly rotate smoothly, but sometimes jump instantaneously.
But watching such quantum jumps between life and death for Schrödinger’s cat would require that we be able to measure precisely an impractically large number of quantum states. For a 'cat' containing 1020 quantum particles, say, we would need to be able to tell the difference between 1010 states – too many to be feasible.
Still, I wouldn't put it past some experimentalists coming up with an ingenious way to test this.
 J. Kofler and C. Brukner, Phys. Rev. Lett. v.99, p.180403 (2007).