I think it's important to maintain a distinction between the syntax needed to actively do math, and that needed to passively visualize it. Of course these types of visualizations are great, but people aren't going to be solving differential equations on their iPads anytime soon...
It's a tall drink of water to say you're going to 'kill math' if really, all that is happening is a divergence from the so-called 'interface'. I think that interface is vastly more relevant than visualizations when it comes to understanding how to think an equation through, and just as importantly understand why arriving at a certain answer matters.
I'm not sure I'm too thrilled about this idea, but perhaps I'm misunderstanding it.
Also: From the ideator's own blog (http://worrydream.com/KillMath/):
analyzing a differential equation without exploring it in phase space was like analyzing a piece of sheet music without actually hearing it.
Terrible analogy, since most musicians do this all the time. It's called sight reading, and it's a very valuable skill to have. (citation: former concert trombone player)