Ironically I agree with just about everything T-C-L said, in principle at the least. It's the things he doesn't mention, particularly the demonstrable fact that Opera has more trouble rendering a wide range of popular sites properly than FF and IE do, that really put me off. Along with the unwillingness to implement certain features that I consider basic and very important, like auto-complete as has been mentioned. Judging by the quality of some of their other features I think Opera's devs could do a knock-out Roboform-type implementation natively in Opera; if they added import for Roboform settings to it I think they'd have a lot of converts based on that alone. Who would bother with an addon/plugin when you could get it natively in an already highly capable browser?
The bottom line with Opera is I'm left to wonder just what their goals are, who their target market is. And here is where it might be better if I was more familiar with the Opera community and developer intentions. I honestly *don't* know what they're aiming for. For all I know they may prefer serving a niche, implementing cool new features that a certain target market is interested in. It seems clear *to me* that they are not *as* interested in courting the mass market as one might think; not as interested as say Mozilla is. If that's really the case perhaps Opera and I are doomed to a tragic romance, where I see so much potential and so many wonderful features, but simply can't deal with the few but key missing features and rendering issues. The sad thing is these are things the devs should be able to easily fix yet they focus on other things like the really IMO irrelevant Bittorrent support. You have to wonder why.
For now I continue to use Opera as my primary browser - it's still a better experience than FF for me - but I continue to be on the lookout for a replacement. Maybe Maxthon...