Check the polar plots at: http://www.acousticsfirst.com/artdif...F_Art_E_PP.pdf
The low end is "always" diffuse (due to size) - while the reflection gets more and more specular as the frequencies gets higher. Specular being a reflection going out with the same angle as it originated. The angle is zero in the case of the plots linked above so we should expect the diffusor to not have energy concentrated in the middle of the plot. Ideal curves for these pictures, for a diffusor claimed to work from 125 to 16000 Hz, would be the 125Hz curve, at all frequencies. As you can see, that's clearly not the case.
Here's some other plots for comparison: http://www.rpginc.com/download/Broch...ping%20Web.pdf http://www.auralex.com/testdata/PolarPlotTFusor.gif
PS: Sasi-sidi: I think the 2D can be better for smaller spaces than the 1D. From what I've understood, the 1D's needs a bit more distance for the scattered waves to mix properly.