I think "Withdraw" in this instance is any movement in combat, period. It lists retreat as a act distinct from moving to a new position in combat. I'm interpreting it as similar to an attack of opportunity. Enemies move past the character more than one square, and they are technically readjusting in combat, and deserve a whack.
I think you will find the relevant passage on page 314: "The judge is always right. Let the rules bend to you not the other way around."
The general rule relates to withdrawal (moving away), but I agree that it should relate to "move throughs" as well. The question then becomes, at what point are there too many creatures moving by you to allow you to attack them all? And that, my friend, is up to the Judge. In my case, I deem that (1) the size and number of creatures and (2) the experience of the character are both relevant.
Imagine that you open a door, and five mid-sized dogs come out. Even if those dogs all have collars and leashes, is it conceivable that you can grab all five dogs? Your interpretation of the rules would make the expected outcome that at least 2-3 of the dogs are successfully grabbed (attacked). My experience suggests otherwise. The literature that inspired the game suggests otherwise as well -- even Conan does not get unlimited attacks on every Pict that moves past him.
Either you can have a ruleset in which the writer must imagine all possibilities beforehand, or you may have a ruleset that gives general guidelines and relies upon human moderation. Joseph Goodman, thankfully, decided upon the second option. So, as with Gygaxian D&D before it, DCC RPG doesn't grant a "court of appeals" from Judge/DM decisions, although (again, as with Gygaxian D&D before it) it does grant the ability to address the Judge/DM not only on specific rulings, but also on whether or not specific rules should be applied to your character under special circumstances.