2024/02/27 DPA: "Poof! Defender Be Gone!"

White to move.

Black's Queen defends the Knight which defends the Bishop.  There's probably something there to exploit.

The bigger threat is Bxg6#.  So let's Remove the Defender:  1. Qxf4 Qxf4  2. Bxg6#.

So Black doesn't have to accept the Queen sac:  he could defend with 1. ... Bd4+, which frees up g7.  But 2. Qxd4 threatens 3. Rh8# so Black must trade and White is up an entire Rook.

The puzzle creator chose 1. ... Qxf4 to end things quickly.  Yes, 1. ... Bd4+ [or, even better, 1. ... Be5] is "better" [ie "less bad"] but the advantage to White is ginormous.

The key was that the Knight was the only thing guarding against checkmate so White happily offers his Queen to get rid of the Knight. 

If you were reluctant to give up your Queen, recognize that checkmate is infinitely better than holding on to your Queen.  After all, the Queen, as powerful as it is, is just material, and material is just another variable that you have to weigh when considering a move/tactic/strategy.

The fact that the move chosen was not optimal does not mean the puzzle was "wrong" or that the opponent "blundered":  the puzzle was meant to show a tactic [in this case, "Remove The Defender"] and it did to successfully.

https://www.chess.com/blog/EnPassantFork/no-the-puzzle-is-not-wrong

Another way of looking at these types of puzzles [that contain sub-optimal opponent moves] is as lessons, designed to teach something.  Viewed that way, the puzzle was a success.

BTW:  "blunder" has a specific meaning of altering the outcome of the game [ie a win becomes a draw or a loss].  1. ... Qxf4 was not a blunder because it didn't alter the outcome of White winning.  It just accelerated the win.