What's also interesting about this is that Israel are working with Pfizer in terms of providing them with all the clinical data on their vaccination programme (including when people develop antibodies on what date if they are in the test group) in exchange for priority access to the vaccine. This might also be a risky strategy if there was some major flaw in the vaccine, but now looks like an extremely good strategy because Pfizer gets the real world data and Israel gets enough vaccine to adhere to the schedule properly.
Three weeks may not have been tested up against two weeks, but the early data from Israel in that article suggests that it would not be good enough at two weeks (15 days, 33% less positivity, I guess that means cases compared with the control?) and doesn't map onto the better efficacy shown in the original trials. I also saw another article, which I can't find now, that shows that antibodies go crazy high after the second dose is given at three weeks, suggesting Pfizer had very good reason to select 3 weeks plus as their point for redosing, although of course, 4 or 6 might equally be great we just don't know.
I'm also guessing the UK isn't going to have the capacity to collect this type of data and so won't know what's going on, especially if they give the second dose at variable times (up to and beyond 12 weeks) and also if they mix vaccines due to lack of availability, perhaps data will be collected on the Oxford vaccine roll-out (she said hopefully).