Can somebody explain me this?

I believe that most book moves here are based on showing up a minimum number of times in the Master database.

It is actually not so simple. Even though the move isn't objectively best, lots of people would stand by it as I have just learned. I do understand arguments that e4 e5 f4 isn't objectively best opening and gives black advantage but to me for most players Qf3 is going to end up being a real bad choice.

But for beginners this is disaster if say f3 e5 g4 and g4 would still be a book move because it was played in tournaments. I would lose my confidence in the whole review thing if I were new. By repeating this kind of 'theory' I would be losing majority of games without realising g4 albeit book move, is actually a blunder!

I use game reviews only for youtube to get that nice mark notation on the board and accuracy for both players but other than that? I would shy away from it. Dont get me started on the comments after each move suggested!

That makes sense again. But lets say masters played some inferior lines for years and engines know that. How can you override this to warn the players who just look up the analysis and move on? They are bound to repeat their mistakes and wont learn

Book has never necessarily meant the good or best. I know there's ongoing improvements with the way classifications are done and I believe work around what is or isn't considered Book.

The reason I was asking is simply because I thought book should be the best 2-3 lines for white and book should end when white will mess up and in relation to this some reasonable replies by black.

Qf3 here, I cant help it, will bring most players absolute grief in the long term should they find nothing wrong with it. Most masters have dropped it anyway as it doesnt score well.