What? What? The claim that land based telescopes are matching or surpassing Hubble is based on the success of sophisticated techniques for overcoming atmospheric distortion.
I think the argument remains largely economic. You can support some number N of these ground instruments for the cost of 1 Hubble, and the scientific return is judged by some measure to be equivalent.
Still, the Hubble enjoys unique advantages, and it would be a great mistake not to exploit them to the fullest, IMHO. It’s just about the best thing we’ve got going in LEO, and if it’s too dangerous to service it, well, let’s just go back to bed.
I've always been a great supporter of the Hubble; actually got to see the mirror and touch the launch package.. Man, am I really that old now? Anyway, the conceptional problem is thinking that we only need six or so launch vehicles, even if they are reusable. We don't need that, we need thirty or sixty. The cost per vehicle, and their value, means the loss of one impacts missions for years to come. And we are talking rocket science here - bad things will happen. You have a controlled explosion of highly combustible materials that flings something into space. Even if you have a ninety-nine percent success rate, you've got a critical failure to contend with every eight years or so.
We don't need a new Hubble, we need three of them, designed to have every component replaced at whim. And making three of them with less shielding, less worry on the ground about parts failure, would in the long run mean far better science and far cheaper costs.