Let's do the math. I'm not particularly great with statistics, but these stand out.
In May-July 2006, Burnham et al. surveyed Iraqi households to measure the civilian cost of the Iraq war. They compared the death rate since March 2003 (during the war and occupation) to the death rate before March 2003 (during the sanctions); the excess had reached about 655,000 +/-130,000 (to 1SD) dead by May-July 2006. In June 2006, the Iraq Body Count had confirmed about 41,000 war-related dead. That implies 15 unreported war-related deaths for each reported war-related death. At present, the IBC has confirmed 62,570-68,593 war-related dead. At the same ratio of 15 unreported war-related deaths for each reported war-related death, this implies 1,001,000-1,098,000 war-related dead.
So when people say that 3,000 people have died in the war, or 68,000 people have died in the war, remind them that 1,000,000 people have died already, and more will die until the war ends. Those who would continue the war until January 2009 will kill 700,000 more people; those who would continue the war until January 2013 will kill 2,400,000 more people, unless the death rate changes, and it is more likely to rise than fall.