Yes, from a problemist's perspective it cannot be proven that black's last move was indeed g7-g5. It could have been g6-g5. But on GK...? I'm not sure.
I composed this problem for prove the posibilities of en passant capture in a puzzle. It is not as orthodox as it should be, But I defended other idea, with an interesting clash between two bishops. In a near future, I'll compose a complete orthodoxe en passant retrograde chess-puzzle. Thanks by the comments.