> ""almost certainly an underestimate", because a number of climate-health relationships are not measurable due to lack of adequate data".

How can the very lack of evidence and data logically support a claim that any estimate is either too low or too high with any "certainty"?