Its only flipping a coin, or 52%, if you completely ignore the fact that the 52% was an average of 0% for the first few days and 90% after a couple of weeks.

Considering that we know and expect that it will take a few days for the vaccine to start working, and we know and expect that the protection after say 7-8 weeks should be close to the efficacy after 3 weeks, I see no justifiable reason to average those figures together.

The data for the first ~10 days should be discarded as they're before the vaccine kicks in.