Aren't those readings based on a tank without water?
I did some experimenting back when I first got to use a PAR meter. I found that the intensity with water is a bit higher than without water. Eventually I figured out that the air to water interface acts to slightly focus the light, increasing the intensity, plus, for most tanks, you get some light reflected off the front and back glass, which also increases the intensity. After that I started just measuring PAR out in the air.
All we are doing is stating that a light which produces XX PAR at YY distance, in air, will give you low or medium or high, etc. light in a tank that has the light YY distance from the substrate. Of course in a tank of water, even with no plants, the PAR varies a lot from one location to another in the tank.