There's really only two ways to speed up the process:
1) Have a pandemic with a disease even more out of control.than this one.
Yep. If there was something that spread well asymptomatically (i.e. a very high R0 ) and a high kill rate once infected, or disfiguring/crippling. Also, there would have to be no readily available prophylactic nor treatment that can be provided to the masses.
Picture something worse than smallpox on steroids. Something like that would mean the benefit/risk ratio of getting a vaccine out is much different than is the case for covid. Hopefully BARDA and others look into what should be the protocol in that case.
2) Deliberately infect the test subjects to see if it works.
Yep. A good quick challenge study combining Phase 1/2/3 (big enough phase 3 to drop trial arms that are deadend from phase 1 results if those come in before enough data from phase 2/3). I think getting volunteers would be almost impossible (with a really bad virus). If there was treatment that worked, then a challenge study might be done. Instead you might have to just get subjects and placebo group from a hot spot, if there is no effective treatment.
Hopefully quick reuse of MRNA technology once a good spot in the virus is located.
Production as soon as first test starts. If you can get production beyond those needed for trial, allow those who have highest risks to voluntarily assume the risk and get it without waiting for trial results[effectively becoming a separate open trial). MRNA are self limiting as long as they are designed to producing something that replicates (The covid spike alone is not self-replicating)
This all assumes the virus cannot be contained/isolated from the public. (i.e. it has gotten to community spread). Hopefully you can get it(effective vaccine) out before it spreads everywhere, otherwise you get the problem of production speed and supply chain. People are right to say what I just described would be rushed science with significant risk. So it would have to be weighed against what the virus was doing and how fast it spread,.
Thank-god we have yet to have a situation like this. But it is one a country should try to plan for(But not break their economy to get perfect). Covid even with 500,000 deaths is not this. Though we do not know the extent of long-covid.
Covid allowed us to implement MRNA vaccines effectively for the first time.
With any vaccine, you need enough cases to accumulate in the placebo group to test for statistical significance. We were actually helped in this matter that the pandemic did rage so fiercely and with little seasonal variation, so we could reach that trial endpoint much more quickly.
The second option would never pass muster with an institutional review board.
A challenge study would pass muster if the situation was truly dire , and there is an effective treatment and minimal harm occurs if treatment occurs on symptom onset. Of course in the later circumstance there would have to be a reason that treatment as a solution did not mitigate the dire need for getting a vaccine out quick. Even a normal study is a tradeoff of risk to the subjects and the benefit to society (and potential benefit to the subjects).
Okay this is a Disney forum so I will not go too far down this bunny trail. Plus this is all speculation on my part.