The new generation of young players possess today a wide range of tools to improve of which I could only dream about in my starting years. Live broadcasts with comments from grandmasters, countless high quality books, extremely strong engines, online training-facilities... it is for everybody affordable with a minimum budget.
The negative side of the big stream of information is that it became even for top grandmasters impossible to keep track of everything. So you need to filter but how and what depends of a lot of factors. The entertainment value should not be neglected but more relevant is of course the educative part. In the past the rule was first to work on the weak points as that is where the most progression is possible. Today however there are increasingly doubts about this old rule. Something you do good is often also something you like to do. As a consequence it often takes much less time and effort to work on the strong points. Eventually some recent studies proofed a considerable gain in profit by working on the strong points instead of the weak points. A book about the Kings gambit will be very easy to read through for an attacking player and will surely show quickly dividends in practice. However forcing the same attacking player to read a book about the Dutch stonewall could be even damaging the results.
So it is not all bad to specialize to some extend in chess. Nonetheless I won't deny there are dangers too. Playing the man is something which you need to take into account especially when the opponent knows your strong points. The success largely depends how successful you can avoid the disarmament of your strong points. A clash of different styles often caused lively discussions in our rich history of chess. We all remember how Botvinnik lost his fist match against Tal but was able to reverse the tables in the returnmatch. We don't even need to go back in time or look to the very strongest players to witness such interesting duels. I noticed end of last year a game played at the Antwerp Liga between Marcel Van Herck and Robert Schuermans in which Marcel outfoxed Robert by simplifying the position and win soberly the endgame.
I couldn't stop smiling when Robert took a week later already revenge again with the black colour. This time Marcel didn't manage to neutralize the chaos.
Robert is for everybody extremely dangerous if he can obtain his favorite type of attacking chess on the board. I also once lost by not finding the right answers in the complications created by Robert or maybe the reader still remembers my article how to win from a stronger player.
From my own practice I remember a recent special case of successful disarmament. A couple of months ago I managed to draw in Open Gent against the surprising tournament-winner, the Russian grandmaster Alexei Gavrilov, while the biggest part of the game I was a pawn down and temporarily even 2. In my preparations I had noticed that my opponent was very dangerous once he had the initiative by employing tactics which is in below example nicely demonstrated.
In our mutual game I didn't hesitate sacrificing a pawn to avoid the positions in which he excells. I even sacrificed a second one when new threats were popping up.
The positional pawn-sacrifices are maybe objectively not completely correct but in practice it worked. My opponent was very disappointed after the game about the result but I don't think that I didn't merit a half point.
How successful you disarm somebody, influences without a doubt heavily the final result. If you don't succeed to neutralize somebodies strong points several times on a row then there exists the danger of creating an angstgegner or also called black beast. It explains a.f.a.i.k partly why some players have mutual scores which strongly deviate from the prognoses made by elo.