I think this thread suffices. Your results are quite valid. A few missed data points are insignificant and won't skew the results.
Nobody conducts a 100% poll. When looking at national trends the sample size is slightly over 3,000 in a population of 320,000,000.
Don't waste your time and effort.