The actual number is not too important.
But do you think that's an actual breakdown point for the plane? maybe 14Ghz, maybe 20Ghz? 25Ghz?
but there got to be a breakdown point! law of physics. Can't be violated!
I think if a plane can carry an overkilled very high frequency radar maybe like a 30Ghz, which way out of normal tracking radar commonly used. I will not surprised if F-22's RCS will shot up dramatically.
If my understanding of radar is correct, even within the 2GHz to 12GHz range, the RCS isn't a constant. Some frequencies will be more effectively dealt than others, meaning there is a sweet spot in all of this. The F-22 is going to be more effective in dealing the 8 to 12GHz band, which is X-band, because that is the fighter radar most of all. The SAM and AEW radar bands will be less effective and so. In the B-2 its the reverse, the sweet spot should be more on the 2 to 8GHz bands, especially between the 2 to 4GHz range. That's because the mission priority of the B-2 is different.
Pointblank is correct in that the higher the frequency you go, the more energy loss you are faced with as the waves travel through the air, and the higher the energy requirement. S, C and X bands are chosen for their general efficiency and balance per mission requirement in the first place. The K bands are very inefficient for volume search. All this means is that the stealth fighters should not be stupid to get too low on the ground or get into range with something like a SPAAG or Otomatic, you know the AA tanks that carry radars on the turret. AA tanks, and for that matter, CIWS both land and naval often use K band radars for fire control. The Oto Melara gun with the 76mm guided projectile uses a Ka band illuminator. For that matter, the other significance of this is possibly using Ku, K or K bands for gun fire control, missile illuminators and seeker heads.
Anti-stealth radars are focused on low wavelengths, which is what you failed to see. That's the wavelengths lower than 2GHz. They have tremendous search volume and range for their power, but what they lack is resolution. The answer to that problem is the big volume search radar has to que a sharper resolution one, and that's where the answer is on the previous paragraph.
There is a certain limit to where shaping is a factor. If the wavelength is long enough, does not matter what the object is shaped like, it appears as an "object", though it may be difficult or impossible to determine its identity because there is no discriminating features. RAM or composites do not work against long wavelengths, because radar absorbency is size dependent, so the material thickness, honeycomb cell or sandwich, or ferrite particle has to be matched in direct physical size of the radar frequency.
AESA is not the cure all for radars; they offer better resolution, redundancy, and high ECM resistance. Not power.
PB, only with the current generation of AESAs, where an element appears limited up to 10 or 20 watts of power. However, elements with up to 40 watts, or maybe 50 watts of power is already in experimental stage. The problem is the tremendous heat, but Gallium Nitride as the semiconductor material might change all that, ushering a new generation of AESAs. According to the Russians, once the beam reaches 20kw, there is so much power you can't hide from it no matter what you do. That means for an array with 1000 elements you need 20w per element.