The below plots were generated from 1000 single-pion events, using a version of the StiIstDetectorBuilder that treats each IST ladder as one big wafer (as was done for all previous IST geometries).  The plot in the upper right shows the efficiency of adding hits to tracks, on the left non-IST hits and on the right IST hits: the efficiency for IST hits is ~80%, approximately as expected, and suggesting that the problem of hits only being added to wafer 9 of a ladder resulted from attempting to have the tracker deal with each IST wafer individually.  This modified version of StiIstDetectorBuilder is available at ~wleight/StRoot/StiRnD/Ist/.