Advanced stats dont tell the whole story, BUT they are usually a better starting point for assessment than "what my eyes tell me."
Any person watching a game comes in with all sorts of pre-conceived notions of what makes a player productive. When someone is productive in other ways, a person unconsciously chooses whether their beliefs about productivity are wrong, or the statistics. It's much harder for a person to change their beliefs and biases than to just attack the data, so the latter is usually what happens.
It's kinda the same thing as when folks that say "the spurs are boring." People come in with preconceived notions of what "exciting" basketball is (sick dunks, ridiculous stats by superstars, etc) and when the Spurs win with ball movement, team play, and a solid pick and roll game, people insist that the Spurs are boring rather than admit that maybe their concept of "exciting" wasn't right in the first place.
Same thing as when people love the James White!(s) of the world because athleticism is usually seen more favorably than basketball IQ. But I'll take Oberto from the early decade over Stromile Swift anyday.