The issue I think is you're really focused on the top level mathematics but are completely dismissing second order or deeper thinking on the subject.

Yes, from a top level view, you are spending more gold per expansion. But, as mentioned, Dust is exactly equivalent in terms of value to Gold.

If you get more Dust value, you by that same token get more Gold value. Sure it may not "feel good" to disenchant cards...but that's not really relevant. We're talking about mathematical value. If you Dust a card you want from an Expansion then regret it later, you're in the same boat, especially if it's an Epic card. Sure, you have "other options" than crafting the card to get it, but tell me truthfully: are you statistically likely to get the EXACT Epic card you want within 4 packs (400 Gold, or 400 Dust equivalent, the same as crafting it)? The answer is a resounding no. The average statistic for getting AN Epic is 1 per 5 packs; and there's no guarantee it's the one you even want. Taking Descent of Dragons into account, that is a 1 in 5 chance of getting 1 of 27 Epic cards. I'm no master of statistics, but that seems like way, way lower than a 1 in 4 chance of getting the exact Epic you want, making the point about crafting being the "only way" to get a card back pretty much moot.

Go even deeper and Adventure cards, at least in my experience, tended to be stronger on average than Expansion cards. Almost every single card form One Night in Karazhan saw play in a competitively viable deck at one point or another, and many were DECK DEFINING (like Barnes and Silverware Golem).

So they are worth even more than the sum of their Dust, which is in turn guaranteed to be more than the sum of the Gold cost required to get them. Presumably. If all the cards end up sucking, I'll retract this.