Ok. Well, in my vast inventory of non-diatonic scales and modes, this one was nowhere to be found. Which isn't to say it's uncommon, because some players 'make up' new scales or combine two or more to make a very complex one.

The root is E, which makes it easy. There are only four notes used in the solo (E, F, A#, B). Either the four notes make up the whole scales, or he just played these ones exclusively from a bigger scale which is pretty much undeterminable. But there is a third choice here. I see two sets of hidden diminished intervals - E and A#, and F and B. We could say he took two adjacent diminished arpeggios, dropped the thirds and sevenths, and combined all the notes. Confused?

The E diminished arpeggio is made up of E, G, Bb(A#), and Db(C#). The F diminished arpeggio is made up of F, Ab(G#), B, and D. These arpeggios come from either of the two 8-note scales (see my theory section - Erik Hagen). The thirds are G and Ab respectively, and the sevenths are Db and D respectively. So basically, you're left with two b5 power chords - Eb5 and Fb5 (note that the roots aren't Eb and Fb; the fifths are lowered a half step). He took those four notes, which are origianlly from different keys, and threw it all in the mixing pot. So I see it as two sets of b5 power chords a half step apart.

Maybe I took this a little far, but I pieced it together in my head much quicker. There are usually several solutions (if there is such a thing as a solution in music theory). Hell, maybe he just put on the backing riff and improvised until it sounded good. Sometimes it better not to sit down and pre-plan it theory-wise - just empty your head of all knowlegde and play until it sounds good. That is the point of music after all....isn't it?