Let's say that in his next expansion, Donald gives us the following card:
Cartographer Jr.
Action -
Reveal the top five cards of your deck, frob the snatz, and put the rest back on top in any order.
With none of us knowing what "frob the snatz" means, we look it up in the rulebook, and the FAQ clarifies that Cartographer Jr's effect is to reveal the top five cards of your deck, discard the revealed Copper, and put the rest back on top in any order. From this, we conclude that "frob the snatz" probably means "discard the revealed Copper". In a later expansion, we get another card:
Super Cutpurse
Action-Attack -
Each other player reveals his hand and frobs the snatz.
Most of us would assume that this card makes each other player reveal his hand and discard the revealed Copper. But maybe Donald would come along and say that no, in this case, because "frob the snatz" is referring to cards in players' hands rather than the top of their deck, "frob the snatz" actually means to discard the revealed silver. We would have a definition like "frob the snatz" means "discard any Copper revealed from the top of the deck and any Silver revealed from a player's hand".
This would be a valid and consistent definition, that works for both Cartographer Jr. and Super Cutpurse. But it's not a satisfying definition; we expect the much simpler definition to be the correct one, so that we don't have to learn what the words means in all the different possible cases.
This, I think, is the issue that Awaclus has with Haddock's interpretation. Haddock's interpretation of the word "while" means something different for self-referential effects than for non-self-referential effects. It is inherently more complex than the alternative definition, so we would want the simpler definition to be the default unless otherwise specified. You can say that "while this is in play" should default to have a scope that applies "while that card is in play", but why should it? No other cards behave that way; all other cards that set up effects above a line have their scope limited either by their text or other rules, and if they don't, we would assume that the scope is not limited because it is not specified anywhere.
It is not necessarily inconsistent to interpret it that way, but you would need to define "while" with a piecewise definition, so that it does something different depending on the context. The simpler definition should be the default, unless there is a ruling that says otherwise.
So you have two possible interpretations of the hypothetical, obviously terrible text that are both valid (and arguing that either interpretation is invalid probably means that you haven't recognized your assumptions as merely assumptions). One interpretation is impractical, untrackable, and unintuitive to the regular player. The other isn't.
I think we all agree that this is not a pragmatic argument, since the text is certainly never going to appear on an actual card. That it is impractical, untrackable, and unintuitive are all irrelevant to this discussion. The argument is about what we know about the language of Dominion as far as it has been defined up to this point.