A method of determining the number of radiation lengths of material between vertex and the EEMC as a function of position by using pi0s, and isolated towers and preshower 1 appears to work well.  We have looked at 2006 data, 2009 data and 2006 Monte-Carlo so far.
This method  could be very useful in determining the correctness of simulations for 2009 (and later)  by comparing relatively small Data and MC runs before running larger simulations.
Updated Sept 3 with vertex cuts and some statistical info.