Correct for this type of survey which is designed to give a general perception of what is going on. In the area of drug safety (my area of expertise), it is way to small to measure important adverse events.
Actually the threshold is variable depending on the study and coverage. With a survey such as this, the number is not as important as making sure the sample space was stratified. Meaning if 30% of the overall population identify as conservative, 30% of the sample space need to identify as conservative; if 15% of the population is black, then 15% of the sample space must be black; etc.
Collecting a sample space such as this is usually more difficult and time consuming since you need to collect more then what you need. Then put in place an effective but random system of eliminating participants until you reach a stratified state.
I did not look at the study, so I have no idea if this was applied here.
Another great statistical tip, there are three types of averages, the mean, medium and mode. Unless you know which one is being referenced, knowing the average does not tell you much. Also, knowing the mean does not give much either unless you also know the standard deviation, assuming the distribution simulates a standard bell.