>This is only good as long as you can guarantee that you're not infected

Why? All of these measures are probabilistic.

P(infect others) = P(infected) * P(transmission)

Simple masks don't have much impact on P(infected), but substantially reduce P(transmission) under some circumstances. Valved FFP3 don't have much impact on P(transmission) [not zero], but vastly reduce P(infected) in virtually all circumstances. It's not clear to me even under this simplistic analysis that it is not superior. And in an enclosed space, the effect of simple masks on P(transmission) is almost wiped out.

>If your mask gets moisture-saturated, just replace it.

New mask every 2 hours?