I think that if you'll focus on just the HTML5 and new stuff without any consideration for legacy things it's still a big undertaking but not impossible. There are projects focused on new browser engines, for example - Servo.