I've already explained
here why your arguments are faulty. It should be added that a poll uses a sample to estimate the opinion of the entire electorate and the margin of error shows how much this estimate may differ. Because this estimate can in no case be the actual results, it can't be simply assumed that its value at the edge of the margin of error closest to the actual results is the correct value of the estimate; it is equally valid to assume that the correct estimate was on the opposite edge of the margin of error. This is why margin of errors are always two equal and opposite values and why any comparisons must be made with the actual estimate
Furthermore, the margin of error is dependent on the number of respondents. If we assumed your logic, a smaller sample with the same results would be even better since it would have a larger margin of error.
