Average QWERTY typing speed is roughly 50WPM. How much faster is Dvorak? 10%? 50%? 100%?
Let's say that it takes a couple of years of retraining from QWERTY to Dvorak until a typical user can reach a 10% improvement. How much typing will they have to do in order to regain that lost time?
Typing, nowadays, isn't the bottleneck. If we still had pools of typists clattering away to reproduce our words then - yes - a more efficient typing standard could be good. But what we have is good enough and the gains just aren't sufficient.
Having two concurrent standards is a nightmare from a usability point of view. I'm sure you've had the experience of using, say, a train ticket booking system which insists that you use a A-Z keyboard. You are slow and frustrated.
I live in a country which hasn't fully embraced the metric system - so people flit around between metric and imperial to all sorts of confusion.
Should we all learn Esperanto? Should we all switch to 220V electricity? Should we all use decimal time? All of those standards are better - but they're not good enough to warrant the upheaval of switching.