A submarine "swims" in the same way an airplane "flies."
But for some reason, in English anyway, we seem to define "swim" such that, by definition, only an animal can do it. Or at least, it must behave similarly to a swimming animal...such as by flapping fins and the like.
Meanwhile we don't think an airplane doesn't fly simply because it doesn't flap its wings.
The point is that semantics tells you very little about reality. And debating semantics is not particularly interesting, or at least a good way of missing the point.
Walking bicycles? Well, ok. That's much more of a stretch, partly because of the fact that there is a human much more directly involved in the propulsion of a bicycle compared to things with actual engines (airplanes and submarines)
And you might say "well, define 'think' ". In other words, that's not that interesting a question, since it is more about semantics than about reality.
Or you say it as Djykstra did, which I think is pretty clever.
And yes, your example about how words differ in different languages does a good job of explaining how this is just semantics. If a question becomes moot in a different language, it is a good indicator that the question is less about reality than it is about semantics.
Something about that quote always bothers me. I'm not sure what, but if I were to try to express the same fundamental idea, I would probably phrase it as "whether a bicycle can walk".