I agree with what you’re saying for the most part, but for a population the size of Australia with 1000 respondents, a 99% confidence level has a margin of error of 4% which is perfectly acceptable. Unless the survey targeted very specific demographics versus a random sample, it should be very accurate.
Is that blood? No…