It's a long-winded defense of using GM fiat as your resolution mechanic wrapped up with a pretty bow on top. It then attempts to slyly claim that detailed descriptions, player-initiated action, and GM rulings all depend on using GM fiat as your resolution mechanic... which is patently and obviously untrue.
Furthermore, its general thesis that "modern games include mechanics for resolving stuff that old school games didn't" is severely selective in its sampling of reality. As a simple example, OD&D included explicit resolution mechanics for monster morale; 4E doesn't.
In the end, sure. That's correct. But you've got to make a conscious decision if you're not just going to roll dice and look at the numbers--especially if there's a stat on the charcter's sheet that says the character is very good at a task.
"I'm checking for secret doors! I rolled an X. Did I find it.?"
That's the point being made.
But in all editions of D&D -- dating back to OD&D in 1974 -- secret doors have been detected by rolling a die and looking at the result.
HERE, THE PLAYER HAS BASICALLY CALLED FOR A SPOT CHECK. THE GM ROLLS THE SPOT BEHIND HIS SCREEN AND GIVES THE PLAYER NO IDEA OF THE RESULT--AS THE RESULT WILL BE ROLEPLAYED WHEN ANSWERING THE PLAYER'S QUESTIONS.
That's how 3E recommends handling Spot checks and a passive Spot check with no roll at all is the default in 4E, IIRC.
Your examples are actually making my point for me: There is no division between editions here. The Primer is fetishizing OD&D's lack of certain resolution mechanics and holding up the lack of resolution mechanics as this marvelous and wonderful thing... while ignoring all the resolution mechanics OD&D does possess.
In short: I would recommend that no one read the linked treatise. It's a poor premise misleadingly presented. Whatever good advice it contains can be found in more productive forms in many different places.