Hi
Y is luminance, Cr is the red minus luminance, Cb is blue minus luminance. Without going into a lot of unnecessary and confusing detail, YUV, and YIQ are roughly the same thing as YCrCb and the n1:n2:n3 nomenclature refers to the horizontal sample rate of the digital encoding system where n1 is the Y channel, n2 is the U channel and n3 is the V channel, assuming YUV is being sampled and all three of the n's are referenced as multpules of the subcarrier frequency.
So, 4:2:2 means that Y is sampled at four times the subcarrier frequency, and both U and V are sampled at twice the subcarrier frequency.
Now it happens that 4 times the subcarrier equals one sample for every horizontal output pixel, while 2 times subcarrier means one sample for every two pixels output. So, in that case, the luminance is sampled at each pixel and the color difference channels are sampled at every other horizontal pixel.
The sample resolution, or color depth, whatever you want to call it is exactly the same in 4:2:2 and 4:4:4. The difference is spacial sampling, not bit depth.
As for whether or not we will ever see 6:6:6:6 (that last number is the sample rate for the alpha channel), I would have to say that there is currently no need real need for subpixel sampling so we are probably not going to see it anytime soon.
Oh, and one last thing. 4:4:4 is YUV or YIQ or YCrCb, but it is definitely not RGB! RGB is a different color space and changing sample rates is not the same thing as color space conversion.