So the bearing and housing and shaft surface were destroyed in two days from the balancing job!
One of the early technical papers on CSI PeakVue presented in October 2001 stated, “the primary PeakVue parameter which should be used for trending PeakVue measurements is the parameter referred to as Peak-Peak Waveform”. I believe that the bearing was significantly damaged by contamination at the time of balancing and when the grease seal was replaced. The PeakVue waveform was already indicating nearly 8-G acceleration before/after balancing! The fan operates in a hostile environment, so consider better shaft seals and improved monitoring with PeakVue measurements.