Seems a bit weird, it'll be fun to find out what the mechanism for this effect is. It seems strange that 2 doses does nearly nothing, then 3 doses suddenly causes the body to figure out Omicron.

Something interesting is that the Omicron mutation seems to very much be about vaccine escape, natural immunity through prior infection offers some protection (linked paper, figure 2C). It isn't totally out off the question that something about the sort of people who get boosters is skewing these results. Table S1 (last page) suggests there are a lot of healthcare workers in the booster sample, maybe they are also people who's immune system is also getting primed from exposure to real coronavirus.