"By next weekend the Pollsters will be forced to publish the real numbers, and those numbers are gonna show a landslide victory for George W. Bush."
Hugh Hewitt (and I) second your conclusion:
The RealClearPolitics' "average" of Bush's lead in the polls, which has gained wide currency in this year of wild swings, is at 3 points, but as the Harris Survey folks admitted yesterday, the turnout models drive the polls, and these models are guesses built upon voting data from before 9/11 which may be further corrupted by the resistance of subgroups to accurate sampling. I send you back to KerrySpot's collection of data from 2002 races in which the polls were not just wrong, but way wrong. In three Senate races in 2002 --in Minnesota, New Hampshire and Colorado, all states figuring heavily in this year's presidential contest-- pollsters, including Zogby, underestimated the Republican vote by 9, 8, and 8 points respectively. Try adding that correction to the current polling on a state-by-state basis.
http://hughhewitt.com/
For additional insight, read the following commentary by Michael Barone -- EXCELLENT!
http://www.usnews.com/usnews/opinion/baroneweb/mb_041020.htm
I also recommend that you read the article posted at the following thread:
http://www.freerepublic.com/focus/f-news/1252882/posts
Very interesting and encouraging, as usual. THANKS!