I guess if you made clearer what the two references frames are, it would be ok, even without mentioning the equivalence principle at that point. In any case, I enjoyed the approach, and it is funny that using only newtonian physics it is possible to arrive qualitatively at the same expression for the angle deviation for small masses.