I should've gone for a more piece active approach. Alternative variation: 20. Nd3 Bc6 21. Nxb4 Qb7 22. Rac1 Rac8 23. Bc4 gives white some piece activity.

Black has the extra pawn and much the freer game. You'd think White had an Everest to climb here! Having said that, I was half-expecting 20.Rab8, here. I note that Black has a change of mind subsequently...

Having placed the rook on c8, one feels he might as well have left it there, or else exchanged a pair of rooks. At that, I might have been inclined to offer to exchange both pairs of rooks...

Alternative variation: 22. Nxb4 Qf6 23. Nd3 Bc6 24. Qxb8 Rxb8 25. Rxc6

Alternative variation: 23... Bf8 24. Bf1 Rb5 25. Rc1 h6

Alternative variation: 24. Qc1 Qb7 (24... e5 25. Bf1 Qd6) 25. Bf1 Rdc8 26. Nc5

24.Qc1 looks the more 'natural', doubling heavy pieces on the c-file. I can only surmise White wanted to double rooks, but that tends to leave the Q locked out of the play.

Alternative variation: 24... e5 25. Bf1 e4

I think I prefer the alternative variation. This is one time I might have been inclined to allow the doubling of rooks, and see how White set about freeing up his Queen.

"Concede"? I presume you mean 'concede' the file by exchanging at c8...

Indeed!

Alternative variation: 25... Rd8 26. Bf1 [allowing for Re1] Qd6

It is just as well c4 is a safe square for the rook, and all. But the doubled pieces makes it hard for Black to improve his position on the Q-side.

Understandable: the move occurred to me as well. This seems to aim for ...Nc6 and ...Nxa5 if given the chance. ...Bd5 comes into consideration as well. But withdrawing the knight from its d5-outpost at this juncture looks risky.