Stephan Narison agreed to contribute to my blog with the following lines describing the current situation about the gluon condensate. It is a pleasure for me to put them here.
The gluon condensate introduced by Shifman-Vainshtein-Zakharov (SVZ) [see also Zakharov, contribution at Sakurai’s prize 1999: Int. J. Mod. Phys A14 (1999)4865 (here)] within the framework of QCD spectral sum rules (QSSR) plays also an important role in gluodynamics. The original value of obtained by SVZ from charmonium sum rules has been shown by Bell-Bertlmann (BB) to be underestimated by about a factor 2 from their analysis of the non-relativistec version of heavy quark sum rules. The BB result has been confirmed later on from QSSR analyzes of different channels including into hadrons, tau-decay and charmonium by different groups, where the most recent value of has been obtained [see for a review my 2 books (here and here) and the last paper on tau-decay: PLB673(2009)30 (here) ]. However, in order to extract reliably this (small) quantity one should work with sum rule which can properly disentangle its contribution where some possible competing contributions due to perturbative radiative corrections and to quark mass should not appear. This feature may explain some results in the literature. The non-vanishing of the gluon condensate and its positive sign has been seen in the lattice by the Pisa group [A. Di Giacomo, G.C. Rossi, PLB100(1981)481 (here); M. Campostrini, A. Di Giacomo, Y. Gunduc, PLB225(1989)393 (here)] and more recently by P.E. Rakow (here). Its positive sign is expected in a model with a magnetic confinement (H. Nambu), while phenomenologically, its eventual negative value would leave to serious inconsistencies in the QSSR approach. Its negative and non-universal values obtained from some tau-decays analysis can indicate the difficulty to extract its value among the competitive parameters present there where in the tau-decay width the gluon condensate contribution acquires an extra contribution compared to some other non-perturbative contributions. In fact, this peculiar properties have been the most important observation that non-perturbative contributions are small in this observable, then allowing an accurate determination of from tau-decays.