Introduction: I am inclined to think that what was in Molyneux’s mind, because of which he drew that conclusion from those premises, was this: that what affects one’s touch so or so could have affected one’s sight such and such instead of so or so—i.e., that what feels so or so could have looked such and such instead of so or so. Thus, for example, that what standardly feels like a globe could have standardly looked like a cube instead of like a globe, and vice versa. Certainly anyway, if you did think this, it would seem to you plausible to say that if a man hasn’t had the experience of how things that feel so or so look, he can’t tell by sight, i.e., from how a thing looks, how it feels. And plausible, then, to reason, as Molyneux does, that, if a man hasn’t had the experience of how things that feel so or so look, he can’t tell by sight, i.e., from how a thing looks, whether it feels like, and so is, a globe, or whether it feels like, and so is, a cube. The best he can do is guess.
It’s a proposal worth looking at, in any case, whether it is Molyneux’s or not. Or rather, under some interpretations of it, it's a proposal worth looking at.
Comment: Classic article on Molyneux's problem; maybe not the main reading these days, but important background/further reading.