ThirdWizard said:
I think the difference might be that you're assuming a level of competance in theoretical DM using the "low magic" guidelines provided, and I'm not. But, then, it took us three years to figure out how to run a low magic D&D campaign without breaking everything, so I suppose I can't make claims to that competance myself.
Heh, I honestly don't trust my competence level that far and I KNOW I don't have the patience for it.

I know it can be done. I'm pretty sure it can be done with DnD rules, but, I think it take a lot more work than I want to put into it.
JohnSnow, I like what you said. I agree mostly. Although, the benchmarks you put are perhaps off, IMO. I would place it more thusly:
Scarcity (1 being rare and 10 being common): 4-6
Power (1 being weak, and 10 being ultimate): 7-9
Mystery (1 - magic is technology, 10 = Cthulu): 1-3
When you follow the demographics rules, magic gets a WHOLE lot less common. There was a reason I had the king lighting the city. The vast majority of the people couldn't possibly afford 150 gp for a single light. When you take the demographics into account, DnD becomes a lot less magic.
I would point to Cauldron from the Shackled City AP. A well laid out city, lots of clerics and whatnot, but, certainly not highly magic-common. Not rare - certainly not, but not terribly common either.
One of the ironic things about this is, by RC's definition, I run a pretty low magic campaign. No magic shops, no core casters (not my choice, the player's chose not to have one), about 50-75% wealth and no custom magic items. And, honestly, not a really large difference in how the game plays out. Sure, they need to rest more often, but, that's about it.
But, yes, low to high magic is a spectrum. There's no cut off at any given point. It's easier to see the extremes because there's a whole lot in the middle.