I feel that in the final position of the game viewer, (I'm too lazy to read the actual stuff you wrote), White might be able to generate play on the queenside before black does anything serious. Black might get white's h-pawn, but his c-pawn is beyond easy. The problem with the bishop pair is that the two cannot help to put pressure on one weakness.
The h-pawn might even be given up just so that white some breathing room. After that weakness is gone, it might be difficult for black to take the f-pawn, whereas the two rooks will work wonders on the seventh rank, which they will reach by targetting the c-pawn.
Apologies for the long post, read below if the variations are too much.
With black, I play the Two Knights against 1.e4 e5 2.Nf3 Nc6 3.Bc4 because I don't like to defend against the IQP in the main line Giuoco. I've been using the Fritz/Ulvestad variation (3...Nf6 4.Ng5 d5 5.exd5 b5!? 6.Bf1! Nd4) since this surprises some opponents and I've won quite a few games where white didn't find 6.Bf1 or did, but later lost material to a Bb7-Qf3 skewer or so. Using it in a club game yesterday where my opponent followed the main line I used up a lot of time, so I decided to look up the theory afterwards. I found many lines on the internet and in books, but it was surprisingly hard to find the theoretically best line and its evaluation.
After 7.c3 Nxd5 the main possibilities are 8.Ne4 and 8.cxd4. MCO-14 by de Firmian (1999) gives 8.Ne4 Qh4 as the main line, but continues 9.Ng3 Bb7 with a line continuing until move 14 evaluated as '=', while Fritz evaluates this as +3.1 for white. Pinski writes '9...Bb7? is sharp and good - but only in blitz games.' and gives the same line as de Firmian as the refutation, but with the whole game. De Firmian mentions Pinski's main line, 8.cxd4 Qxg5 9.Bxb5+ Kd8 10.0-0 Bb7 11.Qf3 Rb8 only in a note and then gives only '12.Qg3=' (trading the queens and giving back the pawn) which is obviously harmless and not critical (even 'slightly better for black' according to Pinski and -0.3 in Fritz). MCO is thus useless for this variation.
Then NCO 1999, which is supposed to have even more authoritative analysis. Nunn also gives 12.Qg3 as the main line and evaluates the position as '=∞'. I don't see what's so unclear about this position. Why can't the authors just analyze it to a verdict? Nunn does not even mention 8.Ne4 (MCO's main line) but gives two inferior variations in the notes instead, both also ending with unclear signs. NCO is thus equally useless.
Tim Harding in his Kibitzer only says "10...Bb7 (...) gives rise to a very complex position where, in practice, Black has done quite well." and gives an example game from 1908 where white goes wrong with '12.d3?', thus saying nothing about the main line.
As far as I know the Fritz/Ulvestad is not recommended by any repertoire book. Therefore we need specialist literature to say anything about this variation.
Mikhalchischin and Beliavsky in 'The Two Knights Defense' (Batsford 2003) give instead of 12.Qg3 '12.Nc3 exd4 13.Re1 Ne7!' with a line ending as unclear, but Pinski (1999!) says '13.Re1? is best met by 13...Bd6!' when Black is just better' and Fritz seems to agree (-1.04). After '12.dxe5!?' they only give Nb4 or Nc3 for black, lines proven to be inferior by Pinski. In the 8.Ne4 variation, like De Firmian, they give only 8...Qh4 with a complex line ending in '+=', which is understandable since Pinski considers 8...Qh4 'dubious' and 'not providing the comfort of equality for Black'. Pinski thinks 8...Ne6 leads to an equal game but this line is ignored by both M&B and De Firmian. This book therefore seems unsuitable as well.
Since the other sources apparently can't be trusted, following Pinski's main line, we get this:
In my ChessBase Mega Database 2010, the position after 18.Nc3 has only been reached twice, one white win and one draw, in games between players below 2300 elo, so I don't think this gives any indication. Unlike Pinski, Fritz 10 thinks White has the advantage (+1). I tried to find this position on the internet and found this article. The author improves upon Pinski's line with 19.Rab1 and thinks white gets the better endgame by giving the exchange back if needed after black sacrifices on g6. I think he's pretty much right. In this case, the entire Ulvestad/Fritz variation (5...b5) seems to be more or less refuted. But how do I know this analysis is definitive? Where's the real "verdict" on the variation?
TLDR: So it seems that here we have a variation that has existed for over 100 years and arises in a fairly common opening, but that's not analyzed properly in two opening encyclopedias, is faultily analyzed in one specialist publication and assessed incorrectly by another, has no other analysis available and is entirely refuted by the analysis of one blogger. How is this possible?
How would a master who plays 1.e4 e5 2.Nf3 Nc6 3.Bc4 prepare against this line, for example? Is there a clear place to go to for the theoretical verdict? Where should I look for the established main line / evaluation in any opening?