Such a viewpoint would be missing the point. For a well-constructed problem can equal the beauty, and far exceed the purity, of anything one may find in a game. You just need to make the effort not only to solve it but to understand what the composer is trying to do.
Take the diagram position, for example. Composed by J Fulpius in 1974, it is White to play and mate in two.
In solving, it is often a good idea to start by looking at Black's possible moves and in this case it should suggest the composer's intention. Black's pieces are very much tied down to protecting each other or stopping mates.
Any move of the knight on g8 allows Nxf6. The rook on g7 is tied to preventing Bxg6. Similarly the Bc6 (Qd5), Bf6 (Qxe5) and pawns b7 (Qxc6), b6 (Nxc5), c3 (Nd2) and h4 (Nxg3) cannot budge.
That should suggest a theme of zugzwang with a key move having no threat, but forcing one black piece to desert its post. All we need is to provide mates for moves of the Nb2, Rg3 and pawns c5, g6 and g5.
The answer is 1.Qe6! with c4 and g4 met by Rd4 and Rf4 mate, gxh5 by Qf5 mate, Rh3 by Qg4 mate and Nxd3 by Qc4. Seen in this way, the original jumble acquires total harmony.