Before continuing, here, it is worthwhile to mention that White's next move 4.Ng5 was for many years very controversial (if indeed that controversy has yet ended!). Siegbert Tarrasch dismissed it as a "beginner's move", reckoning 4.d4 to be stronger. Other respected masters took the opposite view. But 4.d3 is also very playable, a line that can also be reached from the Bishop's Opening, as in the recent jsteven1 vs ionadowman game (1.e4 e5 2.Bc4 Nf6 3.d3 Nc6 4.Nf3 ...)

I thought that "Fried Liver" was an Anglicised form of "Fegatello". This capture turns out to be very risky, as White now gets a lasting attack. There are "safer" lines, the main one being 5...Na5. Black gives up the pawn - and indeed most of his pawn structure - for quick development and some initiative of his own. Equally interesting alternatives are 5...Nd4 (Fritz) and 5...b5 (Ulvestad). But before departing hence, Black could, a move earlier, simply have ignored White's attack and played 4...Bc5!? the Traxler, or Wilkes-Barre, Variation. All these lines lead to exciting play, testament to the richness of the Two Knights' Defence!

Yes, "Fried Liver" is simply the meaning in English of the Italian "Fegatello". But it is 6. Nxf7 that is the move that starts the attack -- your earlier comment "[this game] does throw open to debate how strong the Fried Liver Attack really is" led me to expect the game to be a Black win, since like all openings named "Attack" the name applies to White's choice of move. Incidentally, 5... Nxd5 also allows the Lolli Attack, 6. d4, which is sometimes considered to be even better for White.

At move 8, possibly stronger for Black is 8...Ncb4. White has then to decide whether to allow the fork on c2 in the interests of attack. The game continuation can be found also in my annotation of the game Polerio vs Domenico, played just over 400 years ago. In the following move, 9.d4 used to be considered the best move. Maybe the text is too slow, allowing the Black king to make a run for it?

But Black still has the extra piece. If she can just get his king into safety and her pieces developed...

Black last move is understandable, maybe, but I think I would have left the knight at its fine central post and supported it by 13...Be6. Possibly an even better idea was 13...g6 with the idea of developing the bishop at f5 but also bringing the other to g7. But at the moment, white's attack seems to have stalled, so Black pretty much has a free hand here. If Black does go wrong, it not so much here, but at her very next move...

Suddenly, White attack begins anew, based on his passer. Black's 14th (14...Bd7?) seemed to offer a target for the e-pawn to hit, and it is not easy to set up a blockade. But what else could Black have played in order to develop her pieces? Maybe 14...g6, opening up the f5-square?

One feels that somehow White hasn't made the best of his chances, though he was probably more-or-less OK until this exchange sacrifice. Maybe he thought he would get a piece back after 24.c5!? That is one of the reasons why some people are still prepared to undergo the tortures of the Fried Liver - White's attack, however strong, does not lead to a forced win. With plenty of attractive possibilities, it is very easy to go wrong. At any rate, with White's resources having been thinned out and Black's position largely centralised and consolidated, the latter seems now to have the edge.

Even worse would have been 24...Bxd6?? 25.Bxe6+!. But there was absolutely nothing wrong with the "obvious" 24...Nd5! (25.Bxd5 cxd5) after which it wouldn't be so very long before Black can exchange on d6 and round up the resulting d-pawn. With an extra rook she would have been well on the way to a win. But now...  

Bad, bad luck Joanne. With 24...Nd5 you could well have won that game. The oversight was obviously that nasty double attack discovered by 26.Be5!