No human can remotely come close to an engine in accuracy anyway, not even Magnus Carlsen, but it probably doesn’t matter unless a player is skilled enough to figure out why the moves that Romanovsky considers good are in fact now considered bad.
It is funny how we consider a book better if an author checks and double checks his or her work with an engine, then cherry picks the examples that seems to prove his or her point, rather than see directly how an author might truly assess a position, select a candidate move, and play from there.
Now if a reader cannot understand why a move regarded as good by Romanovsky in 1960 is now considered bad by Stockfish in 2021 (without parroting an engine line), then that reader would need to consider whether the types of concepts Romanovsky discusses and his method and examples would overall elevate one’s current-level skills.
Since the book remains highly regarded over many decades, it might be assumed that Romanovsky has succeeded in showing good examples of high level chess assessment as performed by top level GM’s before the completion of his book in 1960.
I started reading Chess Middlegame Planning by Peter Romanovsky (English translation by Jimmy Adams). I'm reading the first section about planning and went through the first three games.
I have noticed that while running through the analysis, the annotator flags moves as ? and says they are bad, but according to the engine it is the best move or one of the best moves. Also some of the lines the annotator mentions have flaws. The engine finds mistakes or blunders in some of these lines.
Am I missing something or are these just mistakes becoming evident because of engines?