Agreed! Research for a game like DF is stupid because this was back before the scientific method.
Anyway, it's all moot because DF isn't focused on technology. The dwarfs already have the most awesome technology in the land: processing adamantine.
Except that the scientific method is itself a technology that was developed (and refined over time)
Adamantium is the perfect example of why we need SOME method of controlling technology. Doesn't it seem silly to you that the only way to know how to process adamantine is to be raised in a dwarf civ? (Not inborn knowledge, not shared cultural knowledge, just some magic property of the dwarven civ as if the civilization you are a part of has an existence unto itself and apart from it's members.
This is why I think technological advancement should be done by "Great people" or rather individuals making individual discoveries. Then society and/or Rulers can chose to use that discovery. If they don't it becomes either esoteric or lost.
Change 'choose to use' to 'have a viable use for the discovery when it is made' and I love it. Aztecs didn't not have the wheel because they were stupid, they didn't have it because they lived on mountains.
I'm thinking about smelting techniques here. Maybe the ability to build different tools out of different materials for slight efficiency gains. If you only build 1 or 2 of them, they aren't worthwhile, but building 50 nickel pliers could improve copper efficiency from 1 ore:1 bar to 1 ore: 1.2 bar, at which point using the copper screwdrivers to improve nickel plier production becomes a viable positive feedback loop. Until then, the wear on copper screwdrivers costs more than the net copper increase by pliers.
That means that people rich in nickel will use pliers early, but nickel poor societies won't use the advance (even if they know it) until their copper techniques catch up to make it worthwhile.
It's complicated and a pain, and maybe not worth it for the 20-30% metal gains you're likely to see, but it's an extra layer of challenge.
Incrimental improvements to steel are also possible. (basically, we could have a few dozen alloys at varying usefulness all teched out, so what you use is based on what you know AND what raw materials you've got.)
Cold smelting is another efficiency gain. Allow 1.2 bars per fuel unit. Magmaless societies would eat this up, but magma forts wouldn't even care.