The problem I have with the Selke trophy isn't that it's inaccurate. The issue is it's inconsistent.
The voters keep changing what they're voting for. When the trophy was first awarded, it usually went to purely defensive forwards (Gainey, Jarvis, Ramsey, Kasper, Carbonneau). There were a lot of excellent two-way forwards, but they didn't get much Selke consideration because they scored too much (Trottier, Messier, Kurri).
For a period of time, the voters started giving the trophy to strong two-way forwards (Francis, Gilmour, Fedorov, Yzerman) - even though they weren't purely defensive players. The pendulum keeps swinging back and forth. During much of the Dead Puck Era, most of the Selke winners were pretty clearly focused on defensive responsibilities (even if some of them had some scoring talent - Lehtinen, Peca, Draper, Madden).
Nowadays, the voters seem to be looking at two-way play (Bergeron, Datsyuk, Kopitar, Toews). From the 2005 lockout onwards, every Selke trophy winner has scored at least 50+ points (except for Toews in 2013, but he was on pace for around 70 points). In contrast, during the first 15 years the Selke was awarded, only five of those players score 50+ points. Twenty years ago, it would have been inconceivable that a Selke finalist didn't play much on the penalty kill and had close to 60% offensive zone starts at ES.
It's much harder to evaluate defensive player through statistics, so reputation and "the narrative" that tends to form late in the season play a big role.