No it's not, it is basically excuses for poor data collection and assumes that a large number of people are forgetting for they voted for in Dec 2019. In fact the sample is even worse than I thought as it's only based on 300 people and not the 500 it claims in it's headline figure.
As I said...