Perhaps you don't appreciate how a review looking at 100 different setups & methods and 93 of those all arriving at the same endpoint (RF leads to oxidative stress) represents FAR stronger evidence than a single study showing the opposite under very specific conditions.