The overall idea is to give the option to use higher or lower level mats for crafting than what are called for by the recipe. These would allow for greater variety in crafting but would come with the risk of failure during crafting.
I will use blacksmithing for examples since it fits best with the idea, but this could be applied to any profession.
Let's look at Shining Silver Breastplate. It has a very nice stat layout, great for plate DPS leveling. But alas, you're short on iron and you couldn't wear it yet anyway. What to do? Try using bronze to craft it. You'll end up with worse stats, and you might fail, but you'll get a product you can use. Or maybe you're a high crafter trying to get business from a lowbie.
We could work in the other direction too. Again, you see a low chest with good stat layout, but it's so low that it's not useful anymore. Try using higher mats such as mithril. It might fail, but if it works, you'll have a better chest than you might get from quests or instances. And, if it does work, you'll gain more skill than you would from crafting an item normally. This might mean orange chances for a green craft or two points for an orange craft.
Additional mats could also be mixed in to add some additional properties, though also at risk of failure and with some tradeoff of stats. Mix in air for greater agility at a loss of armor. Fire might add a reactive proc, which might hit you instead. Water would add mana and health regen, but the rusting would lower agility. Earth would add strength and armor at the cost of move speed. Shadow, oh who knows what horrible stuff it might do. Almost anything could be added to a craft to change the result, but admittedly the majority would have negative effects (don't add deeprock salt to your copper bars) with only a few percent of additives even giving a tradeoff of stats. Sometimes the tradeoffs would be situationally useful.
I want to see crafting while leveling. However failure would increase the chance of successful crafting, and the chance and size of skillup from the crafting. Failed armor pieces would range from having vastly inferior stats (white quality) to trash, but failed pieces could be deconstructed to recover mats. Resmelt metals, unstitch cloth and leather. The idea is to make crafting not flood the player with items which are mostly useless except for the vendor and so it feels less spammy. In practice you'll still be trying to make a dozen copper chests, but you won't be making a dozen; many will fail, as you'd expect when learning. The mat recovery and skill modifiers are so that failure is a learning method rather than just a waste of mats.
This idea of failure being another path of progression will come up in the quest post as well. I am surprised that WoW has managed to do so much to make failure less painful, but so rarely does it add any benefits. It wasn't until the tri-bug fight in AQ40 that "doing it wrong" had any significance besides death and even then, it was only with hardmodes in Ulduar that 'failure' became something significant: "Who talked to the defense projection!?"
So Amazon is Making an MMORPG
14 hours ago