You're right. The point he is making is that there is no need to maintain a rendering engine if it is no longer a point of differentiation when you can focus your efforts on improving the features that make your browser stand out.
I actually see there being a lot more room for improvement in DOM performance than in JavaScript. Most of the time, poor DOM performance is a bottleneck in my code long before poor JavaScript performance, now that JS VMs are so fast.
> But will developers/users really flock to Opera just because it has a fast JIT? For me, Opera is just something to check for compliance.
I think it's the reverse: users will like Opera if it has UI features they like. It's highly unlikely that Opera would be able to beat Chrome / Safari on speed but there's plenty of room for innovation in other areas now that the web is becoming fast enough for most users.
Not having to support a modern browser engine just be able to start those features is the major win for Opera because they can focus on things users actually care about rather than obscure feature compatibility or baseline competitive performance.
