Amarande: It just seems like this opening has kinda petered out under the weight of analysis, like draughts or tic-tac-toe, even.
When it was less well known you got classics like Steinitz vs. von Bardeleben (7 ... d5? involves Black in difficulties), or the various ways in which the Moller can turn bad in one misstep.
Oh gods, the Moller. Often the wrong move actually looks even more logical than the right one here. Tarraschian-school players probably tear their hair out when faced with it.
E.g., in the game given at the end of this post, where 11 ... f5! is in fact the <coup juste>, while 11 ... Ncd6?, which on the surface looks much better, is in fact a well-known trap where Black is already lost after 12 Qxg7!
Or in the alternative Moller line with 9 ... Bf6 10 Re1 Ne7 11 Rxe4 d6 12 Bg5 Bxg5 13 Nxg5, it seems very rational to play 13 ... Bf5? as Michell did in the famous 1907 cable match game, yet 14 Qf3! is in fact already winning, with Black having nothing better than to play 14 ... O-O and give up N+B for R with a +3.07 advantage for White (Michell played 14 ... Qd7 which lost sparklingly to the deflection 15 Bb5! etc.). Stockfish suggests 13 ... h6 with apparently a small advantage for Black; 13 ... O-O also being alright despite seeming like a castling-into-it situation (it's not, because Bf5 becomes safe after castling. Move order matters!).
But if you're wise to the tactical traps, sadly, this once-beloved line of the Romantic Era, while certainly sound, has really run out of powder. If you still want to get something interesting out of the Giuoco these days then I'm going to pretty much have to suggest you venture into the Evans.
I'll conclude with a short boring example I recently played as Black :)
1 e4 e5 2 Nf3 Nf6 3 Bc4 Bc5 4 c3 Nf6 5 d4 exd4 6 cxd4 Bb4+ 7 Nc3 Nxe4 8 O-O Bxc3 9 d5 Ne5 10 bxc3 Nxc4 11 Qd4 f5 12 Qxc4 d6 13 Re1 O-O 14 Nd4 Qh4 15 g3 Qh3 16 Qxc7 Rf7 17 Qd8+ Rf8 18 Qc7 - and it was already quite past time to sign the peace treaty. 1/2-1/2