Unless you get 100% of infections tracked, the risk from COVID will always be measured to be higher than the true risk. That part is not speculation.

Also, if a large study measured no effect, there likely is no effect, even if troves of smaller studies disagree. Publication bias.