And an user sent me an interesting email with further comments. I'm posting this with his permission since he said was too lazy to register in the forum
I found his email very useful and I'm thinking to follow his suggestion when I'll do the crafting system for the upcoming games
Important question to be asked when talking about ingredients is "Where do they come from?"
Typical answers are:
I Fighting monsters
IV Quest rewards
I and II lead to the next question: How much grind is involved?
Obviously, grinding out 1000 Meteorite Ore is a game in itself, but it's a very lazy, cheap and poor design decision.
Best resource extraction should be tied to meaningful player choice, not grind.
1.1) "You get some, you lose some": For an example, after you have defeated a Dragon, you can either let him go and be taught of Dragon
Magic or slay him and extract Dragon Blood for crafting. Similarly: spend time for optional hunt, be late for something
1.2) Optional tough fight: You need troll blood, so go fight some trolls, who are the creatures you typically want to evade, as they are
hardly defeated without spending lots of healing potions. Bonus points if creature in question can drain your stats or are unpleasant in some
other meaningful way.
Buying -> to justify player spending his cash on ingredients for weapon rather than weapon itself, it has to provide some benefits,
but, preferably, not be a no-brainer (e. g. always cheaper and better). Few additional points of damage doesn't feel so good, so
crafted items better have some special abilities that makes player go "yay!" and reward him for the effort (but there has to be an effort).
Quest rewards -> generally not very good mechanism to use for ingredients often, as players typically tend to do all quests they
can, which will mean they just get random ingredients they don't really need, and not get the ones they actively pursue.
It's a bad idea to have crafting use up the same skill points that primary skills (e. g. combat) do. They belong to a group of "pleasant
buffs", of which you pick the ones that suit your playstyle. Even if weapon skills are nerfed and assisting skills buffed, player will have
hard time sacrificing obviously useful combat skills in favour of "maybe worth it" secondary ones.
Also if crafting system is developed well enough and crafting things takes enough effort from player, crafting skills as prerequisite
should be removed entirely, so that player does not miss the important chunk of the game; instead, such skills could provide quality
increase/price reduction benefits.
3) Crafting process
Temperature, technology, quantities and similar things better be kept away from the player reach, for the simple reason of having too much
combinations, and "trial & error" gameplay mechanics being one of the least enjoyable of all. They provide no substantial benefit apart from
questionable value of "Yay, it's like actually mixing stuff!", but provide no meaningful choices and interesting gameplay. Better system
would be to provide a generic recipe and an array of alterations (possibly represented as "slots"), so that player can pick & mix and
tailor the result to his liking.
Getting a useless/bad item for failed crafting - again, too much combinations, thus a very expesive feature to code. What is the reason
to have failure as an option for crafting? As a rule of thumb, if player crafts something out of ingredients he spent a real effort to
obtain, and fails, he'll either scream with frustration or reload. And if crafting in question is not very important - why have crafting for
that item at all?.. (e. g. converting 100 Red roots into 100 Healing potions is hardly fun)
Minigame-based crafting could work too (see the Food cooking example you've provided), should you aim for a more casual experience.