I definitely agree that terroir exists: all the aspects of the growing environment affect the character of wine grapes. It's pretty easy to show on a fundamental level just by comparing varietal wines from disparate climates made by the same producer. QED. Even different exposures in the same vineyard can be remarkably different. Heck, I bet grapes from one side of a vine vs. the other would have distinct character, especially if one side gets superior sun exposure. That certainly applied to my tomato crop last year where one side of the vine burned in the hot sun, while shaded fruit ripened beautifully. I've also heard anecdotally that growers occasionally harvest one side of a vine before the other due to differing sun exposure.
There are two areas where interpretation of terroir often loses its grounding. The first is attribution of characteristics of the finished wine to specific qualities of the vineyard or growing region. For example, I view Brettanomyces expression as terroir-specific in the sense that the yeast strain likely depends on the region, and the chemistry of the wine derived from the terroir in a given vintage will affect the expression of the yeast. But winemaking choices undeniably are also a factor in this specific example. Linking very specific characteristics directly to the soil composition is dubious at best when the mechanisms are complex. For example, calcareous soil apparently facilitates cation exchange due to the alkalinity of the soil. It's not just a case of the roots taking up calcium cations because they are nearby.
The second issue concerns creation of a hierarchy based on superior terroir. Some people seem to assume there is some "uber-terroir" that intrinsically is superior to others. (Sean Thackrey frames this more forcefully as "viticultural racism.") While there clearly are vineyards and regions that cannot produce balanced grapes (CA Central Valley, large stretches of N. Africa), assigning value to high-quality regions is a pointless task. Much of the coastal CA wine growing terroir is subject to lots of sunshine, large diurnal flux, and low humidity. That produces more fruit driven wines in part I believe due to lower fungal pressure than is the case with continental climates typical in France. That is legitimate terroir expression; whether that is more or less interesting than an earthier, lighter bodied wine is a matter of personal taste.
Ultimately, I think you could substitute "uniqueness" for "terroir expression" when considering whether a wine is worthy. If a wine is not unique, then there is no need to buy a specific cuvee at an elevated price point. It needs some sort of differentiation. When a producer allows some compelling aspect of the grapes and fermentation to show, usually that passes as both "terroir expression" and "uniqueness." It's circular logic that I oppose: "this wine is great because it expresses its superior terroir, and the terroir is superior because it is expressed in great wine." It's a fallacy, pure and simple, and an unnecessary one at that. There are genuine differences expressed in wine, and both aesthetic and economic factors contribute to their valuation. There's no need to invent additional justification to enjoy something for what it is.