I believe a big part of speeding up was parrallelization(spelling?) of the phases rather then the usual running them in series. Also I would imagine that the teams working on these had all the resources they needed, minimal internal delays and no expense spared on budget. I can see how those things combined, coupled with 10 months of “crunch time” work from the teams involved could see a 10x improvement in vaccine production.