Overall I prefer the second-version balance considerably over the previous mix, and I like that it's clearer and more open-sounding than the preview mix too. The guitars make much more sense overall now, although the loss of brightness then gives the mix a slightly dull tonality overall. A general 'smile' EQ curve across the whole mix would probably be sensible -- I was able to do a good job of this with
Variety Of Sound's new Baxter EQ. Even with that, though, I think I'd still try to get some more HF and midrange sustain out of the snare, which is coming across as a bit muffled still, relatively speaking. The guitar texture during the choruses still feels like it needs thickening here too, from a long-term dynamics perspective, although it already works a lot better now than it did.
I wonder whether you might want to make the vocals a little smaller-sounding, though, because they feel like they're undermining the implied power of the backing at the moment. It's a tricky line to tread, this, though -- you want enough vocal to get the song across, but not so much that the backing sounds small. Shaving a little off the low midrange on the vocal can help square that particular circle.
The main thing I'm missing other than that, though, is a sense of size for the choruses in particular -- a hint of reverb space. Not anything with a tail, just something to suggest some wider environment. Maybe just a touch more slapback if you want to keep it raw? I reckon the production as a whole might benefit from a bit more in the way of blend effects too (short reverbs/delays and maybe some subtle widening/modulation), especially during the more stripped-back sections, particularly at the end. You don't want anything that's really 'heard' as such, but just something to glue everything together -- you should only be able to hear it's there when you switch it off!