Yes; 26...Bd4 was much the better option. White can now exploit the half-pin on the bishop to prevent its reaching d4 and establishing itself on a strong square.

Had the Black rook remained on d7, White might have looked into the plan Ra4...; Ra7...: Rxb7 etc. Black's position subsequently might have been very difficult to break down, though.

Yes, but opposite colored bishop endings are notoriously drawn. Usually one pawn is not enough.

The bad part here, is that Black's bishop after ...Bd6; a3 will have very little scope, and no 'anchorage' squares anywhere useful. I would expect the WK to start thinking about invading via the light squares on the Q-side. He may prepare that by crystallizing out the K-side situation first. For instance, if Black can be induced to weakening the light squares on this flank, a second front might be opened for White. On the whole I'd be inclined to leave things as they are, there, and bring the King over to the Queen side at once.

Instead of 29...f6, ...f5 might have been the preferable option. The light squares are dangerously weak right across the board, now. White has been clever with his pawns on the Q-side, too. It would be very tempting for Black to play ...Bd6, but that would be a mistake in view of a4-a5! Bxb4??; a6! ++++++++++++++ Coming back to this game, I'm not so persuaded that 31... Bd6 would be wrong. In fact it should draw quite comfortably: 31...Bd6 32.a5 bxa5 33.bxa5 h5 34.a6 Bc5... To bring the a-pawn further forward, White needs the king, but Black's King would then be at liberty to contest this. So obvious is this to me now, I can't think why I overlooked it in my initial comment.

It looks as though 34...Be7 might just have been good enough to hold. I have a feeling White has gone wrong in this little sequence, but it's hard to figure out how. Quite a few interesting concepts are shown here. Black's bishop looks very bad - almost no scope whatever. Yet the upside of 'bad bishops' reveals itself: it can be very handy in defence. Even though the light squares are weak, White has been left with no avenue of ingress on the K-side UNLESS he is prepared to sacrifice by f4. That does not look promising.

Yes I think Kb8 is the losing mistake and after Be7 black can still hold this.

quite so, and even if white didn't play that, Ke7 immediately is better than Ke6?

I would have thought 43.Ke7 the better option than Ke6, but 43.b5 would have been immediately decisive.

Many thanks! Unlucky! You need to brush up on your endings, but I am sure annotating this would have helped!

This was not such an easy end game for either side. Though Black could have held with best play, the task was made difficult by the lack of scope for the pieces. The one thing in Black's favour was that the plan to invade along the f1 to a6 diagonal was an obvious one such that effective counter measures (...Be7) might have been set in motion betimes.