Is "My Best Games of Chess 1908-1937 by Alekhine" worth?

Volume 1 of the series, from about 1908 to 1922 was one of the first chess books I bought in the late 1970s.  I was a beginner then.  AA analyzes his own games and he is not writing for beginners.  He might hint that the moves played were superior (or inferior) to alternative lines. But he does not spend a lot of time explaining why.  Everything seems to be self-evident which is not the case for anyone relatively inexperienced.  Positional advantage is constant theme in his thinking, and that is a level of subtlety that takes time to understand. 

He seems to be writing for players at about the 1800 level or even higher.  Also, the choices of openings have not been as popular until possibly recently.   

As a beginner, I would steer clear from this book if you're paying full price for a new copy.  If it is a gift to you or you're finding it in a used bookshop for a small handful of dollars, that's different.  By all means pick it up.  You will get something out of it regardless.  

I asked this question because in Silman's book, he recommends the reader to, for example, take one of AA games from his book, cover the opposing players moves, and try to figure out what AA will play next. To analyze what moves would be good, write notes down, etc, then when you have completely exhausted yourself, show what AA actually moved and compare notes with him. Do you think this would be a bad idea? Even if it's meant for 1800+? Thanks!

And I'm sorry, I meant to cover AAs moves, not the opposer

You do not need to buy a book for that exercise which I highly recommend.  Reading chess analysis slowly and carefully from a book like AA's best games is an important part of training.  But you are asking about something else. 

Any chess grandmaster's games is easily downloaded off the internet as a collection of pgn files or other compatible format.  You can download millions of games if you want.  There was an interesting and simple computer program written by Fred Mellender about 10 years ago called Guess the Move.  It has been covered many times here on chess.com.  It is free.   You download it to your computer, load any game into it (as a pgn file.) and play the game, guessing the moves of the winner.  You can even skip the moves in the opening to get into the heart of the position if you like.  The software rates your game compared to Alekhine, or whichever GM you have chosen. and gives you a running score.   There are videos on Youtube that demonstrate how it works. 

Your moves, and the GMs moves are compared to how a chess engine would play the position.   This is a great tool for improving your game.   And I would also suggest you keep a list of candidate moves  e.g. your preferred move, your 2nd and 3rd choices.  That will teach you to think about broader possibilities in chess, and not confine your thinking to one line.  You should hope to have the GM's moves in your candidate list or else you are missing something.  

Again, get the book.  But this software will do that job better and you can apply it to any game which is in a pgn format.  I don't know why Silman would recommend AA per se.  I think many instructors believe there are better GMs to learn chess elements from than Alekhine.  Especially from a beginner/intermediate's perspective.

Let me point out that other chess platforms have developed software which aims to do the same as Mellender's program.  I'm not personally familiar with them.  Once you find something that works, you generally stop looking.  Many would suggest XYZ program is better than "Guess the Move".  But what Silman suggests (covering the text with a piece of paper) worked for him in the 1970s.  We have moved on.  

you are 900 rapid you probably will learn a lot by studying Morphy than Alekhine, there is too much heavy tactics there,  read that book once you reach 1800 rapid