A linear fit to the rate of killing US occupants in Iraq per month since May 2003 (excluding the initial invasion in March and April 2003), using the statistics compiled at http://icasualties.org/oif, shows continued average progress, starting from about 40 US state terrorists killed per month in mid 2003, and now (mid/late 2007) about 85 US state terrorists are killed every month in Iraq.
There were individual high peaks at around 150/month during the massacre of civilians in Falluja ( http://en.wikipedia.org/wiki/Operation_Phantom_Fury ) and following the 2004 US presidential presidential election ( http://en.wikipedia.org/wiki/Operation_Phantom_Fury ) and following the 2004 US presidential presidential election ( link to en.wikipedia.org ) which could be expected to give an average rate roughly constant or decreasing. However, despite these 2004 high peaks, the linear regression fit over the past 54 months gives a positive growth rate.
The number N of US state terrorists killed per month can be estimated as:
N = 36.5 + 11.1 * (y - 2003.25)
where y is the fractional year, e.g. 1 January 2003 is 2003.0, 1 July 2007 is 2007.5, etc.
This fit makes no assumptions about political changes in the US such as the "surge" or withdrawal of the "surge": this fit averages over all 54 months from May 2003 to September 2007.
The prediction for the average number of US state terrorists (so-called "our troops") killed per month by the end of 2008 is:
N = 36.5 + 11.1 * (2008.99 - 2003.25) = 100
Progress? Yes, but not quite in the way "General Betray Us" stated it.