[...] The constituent parts of a combo sell at a higher a la carte price than the combined cost of the combo. This is done to move more product than they otherwise would.
And your loss leader explanation is exactly my point. For example, let's say embersilk bags sell for 300g per unit. Let's say that the current AH price of embersilk cloth is 3g per cloth, but there is so much cloth on the market, and I know my competitors have so much stockpiled, that to move any amount of cloth in quantity, I'll have to keep on undercutting and so will only average about 2.5g per cloth. Let's also say that hypnotic dust is 6g per dust. A naive person might say "The raw mat cost of the bag is 75*3 + 15*6 = 225 + 90 = 315g, therefore I shouldn't craft it". A more wily person might say "In order to extract 3g per cloth, I will sell some of it raw on the AH, and I will package some of it in groups of 75 with 15 hypnotic dust @5g per dust (acting as my loss leader) by crafting embersilk bags and selling it for 300g".
This was a good example... had the topic been solely focused on why people make/sell flasks (etc) below the market price of mats. As I noted in the article though, there could be any number of reasons why people do that, and it is not entirely relevant anyway since we can never really know what price our competition is getting mats at. But that is not really the point of the article - the point is that depending on the volatility of the raw materials, it may never make sense to make any finished goods compared to simply warehousing them.
Let's look at the given Embersilk Bag example, since that is more tangible.
First, it was established that Embersilk Cloth is 3g/each, but will probably need to sell at 2.5g/each to push our supply; Hypnotic Dust at 6g/each, but 5g/each to push. The market price of the final product, Embersilk Bags, is 300g.
Market Price (Buying Mats) = 75*3 + 15*6 = 225 + 90 = 315g
Market Price (Selling Mats) = 75*2.5 + 15*5 = 187.5 + 75 = 262.5g
Market Price (Selling Bag) = 75*3 + 15*5 = 225 + 75 = 300g
So... is this a 37.5g profit or a 15g loss?*
It was never specified what it cost to get the Embersilk Cloth/Dust to begin with, so calculating profit is impossible. It is also debatable how much sense it makes to call something a "market price" when you cannot realistically sell items at that price, but nevermind that for now. My point to you is that if Embersilk is 3g today but 5g tomorrow, that Embersilk Bag better be getting crafted at a 150g profit margin or you were better off holding onto the cloth. Or even if it goes from 3g --> Xg such that you can push your supply at 3g.
Meanwhile, the Embersilk Bag market isn't likely to move from 300g even if it suddenly costs 1000g to make one. We cannot even say "no one would make any new bags at that price" because we don't know how much mats someone may have squirrelled away (and not chosen to sell at the higher prices), or if perhaps the bags have already been created months ago, etc etc.
The Anonymous commenter is absolutely correct that "locking in" Embersilk at 3g each via making bags > selling Embersilk at 2.5g. He is also correct in that selling mats in a package (e.g. any crafted item) allows you to conceivably move more product than individual listings - it is difficult to move Hypnotic Dust at any price on Auchindoun some days, so having the opportunity to "sneak" some into Embersilk is better than not selling any. The issue I wanted to highlight yesterday is this: "locking in" cheap mats into a finished product does not save you gold, it saves you bank space. If crafting something makes economic sense, craft and sell it. Do not craft more than you (believe you) can sell today though, unless you know that the price of your mats will not increase (beyond X% of the profit margin of the crafted good) tomorrow.
There is exactly one scenario in which it might make sense to turn all your mats into crafted items today, rather than crafting only when necessary: Alchemy specializations. It costs 150g to change between Potions, Elixirs/Flasks, and Transmute specs, and it typically is not profitable to, say, make flasks without being Elixir-spec'd - competiters having a 15% competitive advantage over you is massive, and the market itself is generally priced in such a way that the 15% proc bonus is assumed. So, if I wanted to crank out some Mythical Mana potions, it makes the most sense to go ahead and make as many potions as possible after switching specs if I plan on going back to Elixirs afterwards. Transmuting via daily cooldowns is generally profitable enough in of itself that you do not need to be specialized, but... transmuting a bunch of rare gems and/or metagems? You definitely will want to switch.
Incidentally, transmuting rare gems is an example of where locking in prices actually works out most of the time. Uncommon gems typically have extremely low margins compared to their rather massive listing fees such that being able to lock-in a price that works out to be higher than 9g/each has no real downside... provided the rare gem itself doesn't plummet to 15g or something ridiculous.
Hopefully the topic makes a bit more sense now.
*Exploring the concept of paper profit vs paper loss and the realization of both will be the subject of a seperate post.