Early on, we'd have been talking controlled then vs. uncontrolled for almost a year more of unknown future. Given the initial explosion in cases, there was serious "28 Days Later" considerations in possibly sweeping the entire country.
Now we know it's not quite liks that, in speed, but that's with three phases of lockdown (and at least two intense accelerators when unlocked-up again, plus variants that may not be helping) and now at a high enough general prevalence that it isn't burning out either. Like I said, we're now seeing a 'useful' level of community exposure for passive challenges to vaccines, but it's still a patchy test and you're sending out placeboed testers into that same environment.
As you say, we now have a far better idea of how to deal with badly adverse symptoms, so you then seem to agree that a more informative challenge trial is less risky than it was four, six or so months ago (or last April even, when more intense trials could have brung the November authorisations into being a month or two earlier). So you're supporting increased (even if not yet break-even) validity of running such trials.
(Oh yeah, when I talk of volunteers, I mean volunteers. And not in return for the commuting of Life Sentences or anything. I don't know about money (beyond normal Phase II/III practices and what happened with TGN1412) and we don't do things like pay hobos/prisoners for blood donations, etc, so desperation-volunteers are likely weeded out. As-long-as-required support is a given assumption, plus however bad our normal healthcare system is (or has become, under Covid) at least it's not like the US one so famously is .)