Ydnar,
I’m still digging deep into terrains and tcGen and I stumbled upon some hidden information on your website. I saw a conversation you posted that started my headache! 
cloudscapes> well i noticed on your site, your tcGen example, the terrain map x y and z were mapped individually
ydnar> yeah
cloudscapes> but that was one texture
ydnar> that’s pre 2.0.x though
ydnar> you can actually do blends between x-mapped and z-mapped stages
ydnar> have a q3map_tcGen ivector for the first stage
ydnar> and a tcGen vector on the second stage, on a different axis
So, then I found the example cloudscapes was referencing.
http://shaderlab.com/slrl/?p=notes/quake_3/shaders/2001-12-18_vector_01.sl
Is this still relevant with 2.5.2? The purpose as to the need for a different shader per axis is lost on me. Also, and possibly more importantly, can you expand on what you meant by doing blends between x-mapped and z-mapped stages? (maybe an example?) 
(And if I could be so bold to get even more bang out of this thread…)
Also, What’s better to use? ‘tcMod scale’ or ‘tcGen vector’?
Oh. One thing you should know before answering. I’m trying to figure out the most efficient way of blending two textures that are different sizes onto a terrain. (outside of resizing it in photoshop )
a thousand thanks in advance!