The fact that you and your sons have cell phones and VOI, and thus wouldn't be polled, doesn't matter (sorry to burst your bubble). Statistically speaking, pollsters call a few hundred people, possibly at most a couple of thousand, and use these responses to statistically approximate (with MoE depending on sampling size) for the much larger population from which the respondants were sampled. In a country our size, the likelihood of any one person being randomly chosen for participation in a poll is miniscule anywise, so you are practically as likely not to be polled, even if you had a traditional landline - and this wouldn't affect the validity of the poll in the least. The math still works out the same, regardless of whether any one member of the population is sampled.
I should have been more clear. I agree if the pollsters are asking the question are you a Democrat or a Republican, answer - yes and then they ask away....on the other hand, mobile Americans are veering away from traditional phone usage and as a result, how has that skewed the demographic of those who have veered? In other words, are more Democrats than Republicans opting for alternative phone systems or are more Republicans opting for alternative phone systems. That part of a survey cannot capability cannot be ignored any longer.