Define "huge". There are about ~30 mutations on the spike protein from the closest known relative, and there's a fairly strong selective pressure against the wild type (both from vaccine and natural infection). I think it's quite probable that we missed a few months of evolution of a strain circulating in a single-digit percentage of people.

This is particularly true when you consider that this strain has S-gene dropout (Delta does not), and was probably dismissed as Beta and never targeted for sequencing by most labs. That's why I meant by my original comment that we're biased about what we look at.