Cooling the roof externally like that would naturally cool the roofing material, and show up on the flir.
If moisture is simply sitting inside a roof, the temp difference might not be present, the moisture will be same temp as internal air, heating up at same speed with the heat from the sun etc.
The OP needs to try this on a roof that is confirmed to have moisture inside, then perhaps go fly it at some various times.
Eg. flying early morning might see the moisture inside condensated on the roof, and if flown shortly after sunrise, there may be some temp level difference to see in the flir when the sun first hits the roof, and the materials warm up quicker where no moisture is present inside.
Pretty complicated to get right, I feel the M2E flir might not quite be powerful enough for such detection, it might require a much higher spec flir.