I thought ku was the freq. of a rain drop size maybe by going higher it won't be a big problem
Moisture will attenuate, as the frequencies get higher it become more of a problem.
I think both are right. But it depends a bit on how you define attenuation.
If by attenuation, you're referring to absorption of the signal, then, I've seen the water absorption spectrum, and that does increase gradually with frequency, but not by mechanisms like molecular vibration, bending or rotation, or via any mechanism that changes the electronic states of the water, that would cause a significant absorption at specific frequencies. I'm guessing that the increase in effective gain of the dishes (assuming that they are perfect shape) will probably help more than absorption by the vapor will hurt. Think of it in terms of visible light being an even higher frequency, and you can still see through water. Ie there is absorption, and you go deep enough into the water, and it gets dark, but if you're just talking about a general relationship that with higher frequency there is more absorption, that is true, but I don't see it as being that important. In clear water you can see the bottom clearly tens of feet down if the surface is smooth, but if the wind is making the surface choppy, then you can't see even inches into the water.
My opinion is that with Ku, the primary attenuation is not caused by absorption, but is more of a diffraction type affect or even refaction, ie the Ku waves encounter the edges of the droplets and bend around them, changing their direction. I've put a piece of plywood or my hand, blocking the front of a Ku lnbf, and I still get signal, possibly attenuated by 50% or so. However put a bigger piece of that same plywood in front of the dish by a couple feet, and I lose signal completely. My theory is that the plywood changes the direction of the signal, but if right in front of the lnbf, it still goes into the lnbf, but the further away from the lnbf, the more likely the change of direction will cause it to miss the lnbf. Another example of this is that when a storm is approaching, I get Ku rain outages when the storm is maybe 10 miles or so away, typically before it even starts raining at my location, but often get fine reception when it's actually raining AT my location.
Anyway, although absorption does increase with frequency, I really believe that the major effect of rain is often a diffraction type effect, much like a pebbled glass window lets the light through, but doesn't let it go through in a straight line. So if this is the important effect, then I think the size of the droplets and the spacing of them becomes important relative to the wavelength of the signal.
I have one of those WildBlue Ka band internet satellite systems, and I do get rain outages all the time, and it DOES seem to be worse than my Ku satellite, however it is hard to compare. I have noticed times that I'll lose Ku, and not lose the Ka Wild Blue, but more often I've lost Ka, but not Ku. It seems to be a very variable thing, related more to the type of storm and where it is.
The one thing I HAVE noticed is that a heavy FOG will knock out my Ka, but it won't knock out Ku, while several times I've seen it raining heavily, and Ka was still there, but Ku was gone. Ie, it's a complicated thing, and if it was just a frequency related absorption affect, you wouldn't see these differences. It really does seem to be related to the size of the droplets compared to the wavelength.
Just my opinion.
Oh that ought to work real well with rain fade. 
