Yeah, I don't think there's anything wrong with the v2 theme either. I don't think the v3 looks better, just different (it does look new, which has its own appeal).
I know as a creator it can be tempting to change things, to try to improve, but changing for change's sake will anger/confuse a certain percentage of users, while the rest won't really notice.
This happens a lot to me on both front and back end development: wanting to use a new library or framework, and having to resist and tell myself (or have my boss tell me) not to, unless there is a tangible benefit. Pretty is great, and certainly can have a positive psychological benefit, but you just have to be careful not to anger current users in a bid to attract new ones.
I'd say a theme editor is fine, but like jxt said, maybe better would be just to use the same white space as in v2, unless people were complaining that things seemed too cramped, which doesn't appear to be the case.
The UI changes in v3 are great (pausing stopwatch, new tag window, multi-away, etc.), but the new design seems superfluous. It's not that it's bad, just that it's not really an improvement over v2 design. Different, sure, and more trendy, but not better.
As far as unification. Why? What benefits does a unified website/program have? Are the usecases the same? What's wrong with having it all blue? A program isn't a website/magazine/painting and making it look like one at the expense of usability seems foolhardy (though I doubt in this case it introduced any bugs), especially when it requires changing something that already exists.
I know there can be pressure to change or have new things, but look at programs that seem to keep getting more and more popular like sublime text or scrivener or acrobat: their UIs have improved over time but the designs have stayed more or less the same. Compare with the fallout from office, new osx/ios, metro, etc. Change is risky, and doing it simply for "pretty" is both ballsy and reckless.
