Comment: In this paper, we relied upon the Euler angle parameterization of the
4 x 4 density matrices given in math-ph/0202002. However, subsequent analyses
of ours revealed that the ranges of the theta's given in eq. (46) there were
too broad to function as claimed. Therefore, we withdraw the paper. Ongoing
quasi-Monte Carlo analyses (scrambled Halton sequences), with corrected
ranges, now appear to show that if there is an exact simple Bures probability
of separability of two qubits, it is 8/(11 \pi^2) = .0736881