In the 6...Ng6 variation of the Jerome Gambit (1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+) I have always been an advocate of the Queen check "nudge" 7.Qd5+ before picking up Black's Bishop at c5.
The move gives Black one more thing to think about (Where does my King go?) and drives the enemy monarch to the back rank, where it can interfere with his Rook placement.
The following game is an example of the troubles that can arise, otherwise.
Astral1119 - JJayquan
10 0 blitz, lichess.org, 2021
1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Bxf7+
4...Kxf7 5.Nxe5+ Nxe5 6.Qh5+ Ng6 7.Qxc5
I checked with The Database, and found that this direct capture scores 49% for White, while the "nudge" check, first, 7.Qd5+, scores 67%. That's something to think about.
Also, Stockfish 14 shows a clear preference for 7.Qd5+, about 3/4 of a pawn.
7... d6 8.Qe3 Nf6
A couple of posts ago, I referred to this as "A normal Jerome Gambit position" - with the King on e8, that is.
9.Nc3
This is a regular developing move, designed both to protect the e -pawn and prevent 9...d5. However, it fails in the latter case, as Black can play his d-pawn forward, as seen in the earlier game RevvedUp - Crafty 19.19, 2 12 blitz, 2006 (0-1, 14).
In fact, the more modest 9.d3, and the more direct 9.f3 can also be met by 9...d5.
The problem is that White's Queen is on the same file as his King - the one that Black can move his Rook to. As I have cautioned before, "beware the e-file".
9...Re8 10. d3 d5 11.f3
White has reinforced his e-pawn, but in the process he has overlooked something.
11...d4 12.Qd2 dxc3 13.Qxc3 Kg8
Instead, after 13...Qd6, White eventually lost on time in Astral1119 - EliArturKarL, lichess.org, 2021, (0-1, 48)
14.O-O White resigned
White is down two pieces for three pawns.