A perfect game with an imperfect ending

There are a couple of darkly amusing scenes in the movie Death of Stalin, which start with a father being dragged off by the secret police after his son informs on him. Later on (after the death of Stalin) the father is released, which results in a very surprised looking son when dad comes through the door.
Why I bring this up is that when we all get back to over the board chess (which seems to be sooner than I expected), there may be some awkward conversations between former chess friends. I am aware of a number of situations where a surprise win or loss  result has been followed up by a mumbled accusation of outside assistance, or a passive/aggressive congratulations on the quality of play. In a lot of instances the accusations are coming from quite strong players, usually after they have lost to a lower rated opponent (btw this also occurred when players were meeting face to face).
Now one of the causes of these claims is post-game analysis provided by engines, usually on the playing server itself. Some platforms give you an accuracy score (based on the difference between your moves and the engine moves), and anything over 90% is suddenly treated as 'too good to be true'. The problem with using this as proof is that the system is there to see where you went wrong as a player, not where your opponent went right. I've seen plenty of examples where a move has been flagged as a blunder because the best move was +7 while the move played was only +4, while on the other hand, a player has played a perfect sequence of moves, simply because the alternatives are garbage.
In fact in a game I played this evening I manged to get an accuracy rating of 99.2% (while my opponent score 89.9%). Now the weird thing about this game isn't how it was played, but how it finished. After I played 20. ... Bxh3 my opponent replied 21.Kh2? Before I could complete my next move he resigned, which came as a bit of a surprise to me. I was a pawn up and the c pawn was weak, but he still had some play. It was only when I looked at the game a few hours later that I understood his decision. Having playing Kh2 he realised he had left his rook on f1, where my bishop could take it. The only problem with this line of reasoning, was at the moment of resignation I was in the process of retreating the bishop to e6, having missed the capture entirely!


LSmart (1369) - shaunpress (1813) [C47]
Live Chess Chess.com, 10.06.2020