I don't think that qualifies as an undecidable statement. Both can be true at the same time without entailing a contradiction. Think about systems of equations that have multiple solutions. There's no contradiction there.

Well, that's just a fact of the matter about the theorem. The truth of "G" cannot be established from within the theorem. So there's a separation between proof, strictly speaking, and truth.

A statement is decidable if and only it or its negation is either provable or refutable. The fact that the square root of positive numbers can have two possible solutions is provable, therefore the statement is decidable.

A statement is undecidable if and only if it or its negation (exclusive or) is neither provable nor refutable. Namely the Gödelian statement G. To see how G is even stated is fairly complicated; I'd take a look at how Godel arrives at G via Gödel numbering to even begin to appreciate the proof.