In order to see how well our IAA measurement constrains the medium density in models, I started making a comparison following the method by Jamie Nagle. The data used are STAR preliminary, from Ahmed Hamed's analysis with pi0 triggers.

Below, are two plots, one for PQM and one for ZOWW, comparing the charged hadron data (in blue) from PRL 95 (Dan's analysis) and the new preliminary data from Ahmed with pi0 tirggers (in red).

PQM:

ZOWW:

Note that for the charged hadron analysis, the statistical uncertainty dominates, while for the pi0 triggers, the systematic uncertainty dominates. The modified chisquare is calculated using a sum over statistical fluctuations + an overall penalty/scale for the systematic errors. This amounts to assumping that the systematic errors are fully correlated. This leads to problems with the last point from the pi0 data, so I decided to drop the last point, to avoid artifacts in the fit. 

Here is a similar set of plots comparing directly the away-side fragment distribution D(zT) to the ZOWW calculation. The first and the last point of the pi0-hadron measaurement deviate significantly, so I restricted the fit to 0.4<zT<0.8 for both measurements.

Note that while the first point looks OK for IAA, it's some distance off for the D(zT). However, this is 0.3*8 = 2.4 GeV fragments, whichc is arguably on the low side for pQCD+energy loss.