It's an interesting question ... but there's no way I can think of where you can equate absorbed energy per unit area with temperature!
Although energy being absorbed by the body (neonate in this case) is scientifically correct, I doubt that the clinicians think much beyond the temperature reading on the warmer, and simply assume (hope?) that the reality of the output is something similar ("distance" is a factor here, surely).
Why not check out temperature outputs of as many warmers as you can at a specific radiator-to-sensor distance, and see if results fall within a recognisable pattern you can rely on?
