This is a great point. In order to get statistical significance, you need to subject the same number of people to the virus. In a non-challenge trial, we feel

This is a great point. In order to get statistical significance, you need to subject the same number of people to the virus. In a non-challenge trial, we feel better about ourselves because the challenge is "natural", the subjects would have been exposed to it anyhow. In a challenge trial, you deliberately expose them. Either way, roughly the same number of people will be exposed to the virus. Meanwhile, maybe 1,000,000 people a day are catching COVID. The bottleneck that stops that from continuing every single day involves some risk (if the vaccine doesn't work), of a few hundred people, who are in the most favorable position if they catch it, who volunteered. "Better 1,000,000 people die from COVID than one person die that tried to stop it."