IIRC, we'd been told that the Oxford vaccine was most effective with somewhere around an 8-10 week gap.
The trouble is, they were subdividing their groups, rather than having full groups for each dose strategy, so it's in the secondary data analysis, and with correspondingly low confidence intervals (and being not-yet published, and therefore peer-reviewed).
If we waited for this sort of thing, you'd need full sample sizes for each sub division - so single dose, half dose-full dose, full dose-half dose, full dose-full dose for each of 2 week, 4 week, 6 week, 8 week, 10 week and 12 week intervals. So 24 times the number of volunteers (or waiting 24 times as long - so 12 years rather than 6 months).
This is relatively normal in medical trials (though I'm far from an expert in vaccines specifically). You answer your primary researrch question, but design your trial to allow for further, deeper anaylsis of the data collected - but that secondary, with all the problems associated with being secondary.
Incidentally, whilst that stuff hasn't been "peer reviewed" - as in published, discussed, and reproduced as necessary; they're incredibly open about it all in these instances, and sharing their results amongst their peers / SAGE various CMOs etc.