I have done a little study to asses the experimental uncertainty on the mb parameter. There are measurements of lambda_1 and Lambda_bar from the hadron mass moments. The extraction yields a result in which the parameters are correlated to a fair amount (0.7 or so). The error ellipse of these two parameter can be mapped onto the contour of a and m_b. Due to the non-linear mapping this leads a much distorted contour. As a consequence the variation towards smaller values of m_b has to be assumed almost twice as big as the upward variation. http://www.slac.stanford.edu/~elsen/amb.pdf With resolution of a few 100 MeV in m_X we should make a serious attempt of extracting a branching ratio below say 1.55 GeV and quote that as a solid number in the paper which will remain valid in the future. It would be a pity to have the theoretical uncertainty dominate. Eckhard