Polling

Your concerns are well-founded. Major demographic shifts are accompanied by underlying psychographic nuance, unfortunately the implications of these nuances as they pertain to polling are vague. It does seem that certain types of political polling skew Democrat, perhaps because Democrats are capable of articulating a position and are therefore eager to do so, IMHO. But seriously, there does seem to be considerable literature available at libraries on these matters. and it is a matter of considerable debate after the massive exit polling lapses appurtenant to the 2004 election scandal.

I do think you should be careful about rationalizing the behavior of the general public on your own behalf - for instance while 20% of our demographic uses cell phones as a primary point of contact (like you ad I), the pollsters are very capable of controlling for it: "While cell-only voters were more supportive of John Kerry than voters overall, they were similar to other voters within their own age cohort. Because of this, preelection telephone surveys that weighted their data appropriately by age were not significantly biased by the absence of the cell-only voters."

Then again, this doesn't help account for the massive polling lapse of '04, does it? Maybe there result is that our confidence intervals keep narrowing and the margin of error keeps climbing?