There is a long history of trying out different ways to teach physics to undergraduates. In my undergraduate college the faculty experimented with starting with quantum mechanics and ending up with classical physics; introductory calculus was sometimes taught using nonstandard analysis. My thesis advisor in graduate school wrote an intro physics text that included simulations from the beginning, in the `80s. He even wrote a special version of Basic that used vectors to make it easier for students. In Germany it's popular to teach mechanics using momentum currents instead of the force concept.

I agree that introducing numerics early is risky: it's easy to get what looks like a smooth, convergent solution from a naive calculation that happens to be wrong.