They come up with an interesting excuse, but funny that it's a testable hypothesis which they will not bother to test. I have only seen a couple 'meh' studies suggesting infection times are shorter in the vaccinated, which could easily be shown by p-hacking. If there are shorter durations in the length of infection, I would love to see the distributions. I'm guessing there is so much overlap it would be difficult to see anything like this turn up in a secondary metric like antibody responses on secondary infections