Were I inclined to "use up" (euphemism for "waste") time on this problem, I'd look again at how you prove that operations which result in irrational numbers really do.
Take pi, for example, or any of the irrational constants.
In the case of 2 ÷ 3, it can be seen that the next digit "down' will always be 6, and no need to explore that down to the billionth iteration.
I would say the "proof" is that every time we choose a number, the operation yields the same result. Not "rigorous," but good enough for me. Sort of like empirically-derived engineering solutions.
One thing I wonder about, without trying it, is to see what happens if we perform analogous operations on different numbering systems (radices) such as hexadecimal or (egad!) binary.
I certainly understand the desire to develop a rigorous proof of the assertion.
But I'm old enough to not wish to budget more time on it.
As I tell folks, the older you get, the more practical you get.
Terry, 230RN