Going back to the original post for a moment, if I might.
Quasqueton said:
Whenever someone talks about a preference for, or a setting is, low magic, they always comment, "there are no Magic-Walmarts" (or Magimarts, etc.). This kind of statement makes no sense.
It is interesting to note that on one hand, the OP opines that the phrase is widely used, and presumably understands what is meant by the phrase, yet on the other hand that it also makes no sense.
The setting suggested in the core rules has no "Magic-Walmarts". Greyhawk has no Magic-Walmarts. I'm not real familiar with Eberron or Forgotten Realms, but I don't think they have Magic-Walmart-style stores either.
Examination of the core rules and various settings has shown that the "best reading" says they do not mandate Magic-Walmarts. However, it is nonetheless quite easy to read the core rules (and some settings) and conclude that they do have Magic-Walmarts. How these sections read depend very much upon the "filter" applies by the reader. The Magic Item Compendium suggests that, if a player asks you if he can buy any particular magic item, that you say Yes. No gp limit is mentioned. It is very easy to read these rules differently than they were, perhaps, intended.
The only times I've ever heard of anything like a Magic-Walmart in a D&D campaign, it was in a 1984 Dragon magazine, and when I played one game session with a new DM around 1991. Both of those were aberrations from the norm.
Those who believe that Magic-Walmarts are a 3e phenomenon, please take note. In 1984, 1e was the norm. In 1991, the game du jour was 2e.
So, saying your preference/setting is low magic "with no Magic-Walmarts" is like saying your preference/setting is low power -- no god killing PCs. 99% of everyone's preference and setting qualifies as low magic if the definition is "no Magic-Walmarts."
This doesn't follow. Quasqueton offers scant (if any) evidence of what the norm is, and has another thread devoted to the question of why what seemed like the norm in 1e for various people is so widely diverse. There is no reason whatsoever to assume that, merely because Quasqueton's experience includes few Magic-Walmarts, that the average experience is the same. Moreover, we have commentary from several people on this thread that contradict Quasqueton's experience. Even Quasqueton's experience, unless he has played with over 100 DMs, precludes his conclusion that 99% of everyone's preference and setting qualifies as low magic if the definition is "no Magic-Walmarts."
Therefore, we must conclude that this premise is false.
Is "high magic" defined by the existence of Magic-Walmarts? If so, there are very, very few high magic settings. Other than the two strange situations I mention above, I've never seen or heard of any.
While the inclusion of object X might make a setting high magic, it doesn't follow that the inclusion of object X is mandatory to make a setting high magic. While the inclusion of object X might be a component of a high magic setting, it doesn't follow that the inclusion of object X automatically makes a setting high magic.
We can safely say that "high magic" is not defined by the existence of Magic-Walmarts.
Moreover, this is reinforced by the observation that whenever someone talks about a preference for, or a setting is, low magic, they always comment, "there are no Magic-Walmarts" (which must be taken as a bit of hyperbole). If someone equated high magic to the inclusion of Magic-Walmarts, they would not need additional clarification
So why does this phrase and comparison exist as a measuring stick? If a DM was trying to entice me to his game by saying it was low magic because there are no Magic-Walmarts, I'd have laugh. "So, it's just like Forgotten Realms, then?"
Clearly, this phrase and comparison exist as a measuring stick because there is not a strict correlation between low magic and no Magic-Walmarts. One can have high magic without Magic-Walmarts, and one can arguably have low magic with Magic-Walmarts.
RC