Using your analogy, it's more like saying that we know water above 40 degrees will be progressively more likely to cause a burn, so the limit of average heating

Using your analogy, it's more like saying that we know water above 40 degrees will be progressively more likely to cause a burn, so the limit of average heating of a cellphone is 20 degrees. > So let me clarify my original question: is there any evidence of actual harm at the highest SAR values of any cell phone? I don't think there is any.