A recent event I'll describe shows why doing the protocol right away tells you nothing.
I don't want names attached to this, but I had a front row seat for this, very recently. A young man in his early 20s, very athletically fit, is involved in a car accident, and bumps his head. All during the police routine, he is perfectly lucid, if suffering a mild headache. He gets home, and his parents give him a version of concussion protocol - light and sound sensitivity, quizzing on details he should remember, and he passes with flying colors.
Then, two hours or more after the event, he suddenly feels sick, suffers sensitivity, and is taken to hospital. CAT scan is normal, all the other investigations point only to a concussion. He goes home, and begins his recovery.
Point is, the time between when the bump occurred, and when symptoms appeared, was at least two hours. On average, a hockey game is over by then, if it happened on the ice. In this case, he would have sailed through protocol, got back into the game, and faced whatever consequences there might have been.
I would not be surprised if most concussions only show symptoms some time after the event. Even a day later. You see it in the NHL with some regularity. A guy gets hit, returns and finishes the game, and a day or two later, has a confirmed concussion. One I remember vividly was Grabovski. He gets whacked into the boards, is woozy on his feet returning to the bench, later scores a highlight-quality goal on Tim Thomas, and a day or two later, is out with a concussion.
So what is the protocol doing? Satisfying legal concerns? Because it's not picking things up, and is disrupting the game in the process. You can't effectively suspend a player for two days for getting hit on the head, after all.