Official amateur opinion:
I feel like white is worse BUT I think that objectively Black will have a hard time converting.
I feel like Black's Knight will be required for a decisive breakthrough but that it might take some precise move to access white's isolated pawns.
My feeling is that white will come under pressure but might be able to force liquidation of some pieces and that Black might be extract a passed Apawn but that will be it with opposite coloured bishops. And A pawns are always a long way a way from Kings (obv).
But thats just my opinion.

Nb3 and I think White is close to establishing a draw. Qxc4 exchanges queens and frees the bishop; I don't think White will really mind the pawn deficit, especially given opposite colored bishops. Even if Black doesn't take the pawn, I think it's likely (after Nxb3) that White plays c5 anyway, just to be able to put the bishop on a decent square. And in that case, White will control the b-file and be able to target the a-pawn.

I have to say I'm surprised...when I look at this around the time Bxf2+ I thought Black would be able to get more.

My position is gay. GAY GAY. Clearly very little chances for a win. I still have drawing chances but that means this correspondence game could go into a LOOOOOOOOONG grind. Maybe until 2013. anyway originally I wanted to play 26. Nf3 but I wasn't sure about this anymore because of just Qxc4 and my queen is put to the test and if I move it my bishop is attack and my pieces are just loose. Qd3 grabs an open file and maybe I can get to d5. I would like to think I have drawing chances because there are opposite color bishops on the board but the reality is that there are significant heavy pieces left and his bishop has a lot of space while mine is a big pawn. Even if the queens got exchange I wouldn't be surprised if the computer gave something like -0.80 for black.

Not really a move i wanted to make (i wanted the bishop to look after the weak dark K-side squares and reserve the d4-square for the knight) but i convinced myself that i can't allow Qd5 if i want to make anything of the position. Qd4 looks even more drawish, so i'll go with Bd4. Unfortunately, after Nb3 we can probably soon think about shaking hands anyway.
Either i overestimated my position quite a bit or i committed some inaccuracies along the way. Very curious to read the spoilers...still let's play some more moves and see if something turns up.

My pawn cannot be taken on c4 because either his knight or queen will be hanging. If he takes the knight I will play Qxb3 to dominate the file and then plant my bishop on d3 like a big pawn and defend passively until I lose.

I thought Nb3 ideas at moves 26 / 27 had some drawing chances for white, but just Nxb3 from black and white is lost. I'm gonna read the thread to see how white got so defensive and why he thought he'd be ok

No other choice really.

I might be overlooking something major or mis-reading your analysis, but how is white lost here? The position looks very drawish to me. Given the lack of material on the board, static pawn structure, white's control of the b-file, and white's extra pawn (a pawn is a pawn) I don't see any winning chances for black. The only advantage black has right now is a more active bishop, but that's it.

Count again. I think it's the missing f-pawn that gets you. Almost fooled me.

creating some luft for the king before trying to bring my rook into play. Another idea might be to get the pawn majority rolling one day with f5 (although i'll have to be careful not to weaken my K position too much). Probably he will now protect his e-pawn with Bd3 and then invade on the b-file, with a likely queen exchange and a dead drawn position resulting.

You're right, I completely forgot about that. I just assumed it was there for some reason. Oops. Thank you sir.

Bd3, as Noir_Desir anticipates, and I think it's a draw. I don't think White's kingside is weak enough to cause any real problems, and White has the b-file as compensation. Even without the c-pawn I think White is fine here...

22.bxc4 Qb6 Perhaps instead Rfb1

Agree with your assessment about the c-pawn. I don't see how black does much here.

Amazed at some of the evaluations ITT. Black is obviously clearly better here while the queens are on the board, and I can't see why they should ever come off. White's king is weak and his bishop is useless. Be3-f4 looks like a strong manoeuvre soon. Not enough people have studied Dvoretsky's books apparently

Black's position is "better," but what good is the bishop on f4? The position after 1...Be3 2...Bf4 3.Rd1 is still dead drawn and I gave Black two free moves to relocate the bishop.

This is what I am taking issue with, its not about specific moves. I don't see how the position is dead draw at all.

To be honest, the more I look at it, the more I am thinking White is just completely lost...

This is that position: Fair enough, I can agree that dead drawn is a poor choice of words, but I could hold that as White against substantially stronger opposition. To me, Black is only optically better--I don't think there's any continuation which changes that. The maneuver you suggested might even make Black's position worse than the position on the board--and that was with two free moves.

It's hard to find a productive plan for Black: moving the queen allows c5, when White's position is easy to play with the bishop freed, and every exchange moves this closer to a trivially drawn bishop endgame...

I don't know; if you think White is "completely lost," you should be able to give some plan for Black, if not a sequence of moves. That's a far biggest exaggeration here than calling the position a dead draw.

Guess I'll start with Rc8

Ok, in any case there is no hurry to play Bf4 so soon. I would use my second free move to play Rd8, then after Rd1 and the exchange of rooks I reckon the position would be incredibly promising for Black: I can play my bishop to f4, queen to f2, h5-h4, Kg7 and White likely falls into some zugzwang at some point, or f7-f5 might be strong depending what White does. Again, its not really a question of specific moves.

Another decent plan in the original position is Rd8-d6-f6

Black is better. Soon it will either be slightly better or completely winning (10-20 moves)

Have a look at the following game: The position Black won from here was obviously much less promising than the one we have ITT, but nevertheless a 2650 managed to lose it...
[Event "Magistral Barcelona Casino"]
[Site "Barcelona ESP"]
[Date "2006.10.20"]
[Round "2"]
[White "Korneev, Oleg"]
[Black "Bologan, Viorel"]
[Result "0-1"]
[ECO "B17"]
[WhiteElo "2657"]
[BlackElo "2659"]
[Annotator "Alexander Baburin (www.chesstoday.net)"]
[PlyCount "110"]
[EventDate "2006.10.09"]
[EventType "tourn"]
[EventRounds "9"]
[EventCountry "ESP"]
[EventCategory "14"]
[Source "Chess Today"]
[SourceDate "2006.10.21"]

1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nd7 5. c3 Ngf6 6. Nxf6+ Nxf6 7. Bc4 g6 8.
Nf3 Bg7 9. O-O O-O 10. Re1 Qc7 11. Qe2 Re8 12. Bg5 Nd5 13. Rad1 Bf5 14. Bh4
Rad8 15. Bg3 Qc8 16. Bb3 Bg4 17. h3 Bh5 18. Bh4 Bf6 19. Bxd5 Bxh4 20. Bb3 Bxf3
21. Qxf3 Bf6 22. d5 ({Computer suggest the uncompromising} 22. g4 $5 Rf8 23.
Qf4 Qd7 24. g5 Bg7 25. h4) 22... cxd5 23. Rxd5 Rxd5 24. Qxd5 e6 25. Qc4 Qb8 26.
Rd1 b5 27. Qc6 Rd8 28. Rxd8+ Bxd8 {Diagram #While this position looks drawn,
White has practical difficulties - his bishop is "biting at granite", while
his counterpart can be very useful in the attack upon f2.} 29. g3 Bb6 30. Kf1
Qe5 31. Qa8+ Kg7 32. Qf3 b4 33. Qe2 Qf5 34. Kg2 bxc3 35. bxc3 Qc5 36. Qd2 Qc6+
37. f3 Qc7 38. Bd1 Qe5 39. h4 Kg8 40. f4 Qc7 41. Qc2 Qc5 42. Bf3 Qg1+ 43. Kh3
Qe3 44. Bg2 Bc7 45. Qe4 Qxc3 46. Qc6 Qa5 47. a4 Bb6 48. Qe8+ Kg7 49. Qb5 Qc3
50. Qc6 Qa3 51. Qe4 h5 52. Qc4 Qc1 53. Qe4 Qa1 54. Qc2 Bd4 55. Qc4 e5 0-1

Carlsen, Karjakin and a lot of other GMs would keep playing for a win as Black here and even expect to win against a competent opponent sometimes. Of course, White isn't dead lost and "every Russian schoolboy knows" to play on.