I'm going to make some suggestions about how White's positional play could have been improved. Since my rating from 10 games on Gameknot is similar to White's, my analysis can be taken with a pinch of salt, but the game interests me so I can't resist....

1... d5 can lead to the Queen's Gambit, but since White has not yet played d4, an alternative is the Réti Opening, 2. c4.

A better try for advantage against the Stonewall is to play for e4.

Black's next move obtains a firm grip on e4 and renders White's e-pawn quite awkward. On the other hand, it permanently weakens e5, so White should (in the long-term) try to establish a piece there.

That is typical Stonewall variation of Dutch defence. It could be achieved with "standard" order of moves: 1. d4 f5 2. c4 Nf6 3. Nc3 e6 4. g3 Be7 5. Bg2 c6 6. Nf3 0-0 7. 0-0 d5 (or any other substitution).

9.Nxe4 here might not be best, because after 9...fxe4, Black's weakness on e5 becomes difficult to exploit. After 9.Qc2, Black could probably force White to capture on e4 after all, but only by misplacing pieces with moves like Bb4 or Qa5. Otherwise, although it would obstruct the LSB and create a weakness on e3, White has the longer term option of chasing the knight away with f3.

If I were Black, I'd be quite happy with the doubled pawns because they control the centre. The e6 pawn strengthens the pawn chain while the e4 pawn means White must occupy e5 now (leading to the exchange) or never.

True, but it's still a lot better than Black's LSB, so maybe White retains some small advantage? Black's pawn chain could be attacked with f3 at some point, probably forcing exf3, in which case Black's weaknesses on e5 and e6 become good targets again.

It's possible to argue for 13.Bxg5 on the grounds that it removes White's bad bishop (at least, it's bad after 13.e3) and Black's good bishop. On the other hand, White's DSB is actually pretty active even if it is officially "bad", so I'd consider playing 13.e3.

To the annotator's question: Yes. To kicorse: White's DSB is not bad because it is outside the pawn chain. That is the only definition I can accept.

Typical plan for Black is an attack at King wing with support of his pawns and Queen. Bishop at Bf7 blocks Queen route to King wing and pawns of King side rest at their place. They have run forward (with shield or on shield!) For White - all chances at Queen wing...

I agree that White's future is on the Q-side, and Black's on the K-side. I wonder if White shouldn't connect rooks and prepare (and play) a4 and b4 before releasing the pressure on d5?

If either player did have an advantage, they wouldn't be able to cash it in within Fritz's search range, so we're relying on the software's positional strength here. I count two significant imbalances: (1) White's bishop is better than Black's; (2) White has more queenside space. That may not add up to much, but I see nothing for Black, not even kingside play after that exchange.

I can't see what this move does that's positive. My favourite idea here is to attack the e-file weaknesses with f3 (possibly moving the queen first). Alternatively, White could start a Q-side advance immediately.

Black's bishop has really improved a lot, even though it is still "bad". Avoiding the exchange with 19.Be2 looks tempting, but White might be concerned about the kingside threats that emerge (contrary to what I wrote at move 16) after Black plays g5 and Bh3. However, if White is going to leave the bishop where it is, why 19.a3 rather than 19.a4? The breakthrough square looks like b5, and so White needs a pawn on a4. Admittedly, White's queen will also need to be freed from the defence of the bishop so a couple of pieces can support the advance.