When checking for overlaps during construction of a detector geometry, spurious warning messages about "Overlap with mother value" are issued. But from http://geant4.web.cern.ch/geant4/G4UsersDocuments/UsersGuides/ForApplicationDeveloper/html/Detector/geomOverlap.html it is clear that such overlaps are perfectly valid. An "Information" message would be legitimate here. This issue was discussed in http://hypernews.slac.stanford.edu/HyperNews/geant4/get/geometry/980/1.html and I believe their suggestion of offering a verbosity level that disables these messages is a good one, precisely for the reasoning given back then. However, this is independent of the wrong categorization of these messages.
Could you please provide a simple test case which reproduces the problem you're having? There should not be spurious warnings in the way overlaps are computed in the new release; if reported, they're likely real, or we need to understand why false reports are given. It is possible already to turn off verbosity of the overlaps-checking and tune precision in order to exclude messages from tiny overlaps for debug purposes, as reported in the documentation.
