The benefit went up. The acceptable costs of parallelization similarly went up.
(Some steps got easier, too, for the same "everyone really wants a solution yesterday" reasons, like recruiting trial volunteers. For example: https://www.sciencenews.org/article/covid-coronavirus-vaccin...)
> For instance, it took just under 16 weeks to recruit and enroll more than 43,000 volunteers for the final phases of testing Pfizer’s vaccine. When volunteer recruitment began for clinical trials of the rabies mRNA vaccine in 2013, it took 813 days to get 101 participants enrolled. Based on this comparison, that’s roughly 730 days — nearly two years — saved in recruiting alone.
The same article cites precisely one of those parallelization steps I'm talking about:
> In a final jolt of speed, some pharmaceutical companies, bolstered by big vaccine contracts and research cash from the U.S. government’s Operation Warp Speed, were also churning out doses during clinical trials in the hopes that the vaccines would work. Once the companies had emergency use authorization in hand, they were ready to ship doses immediately.