(Now quite, apparently. My position hasn't changed.)
It would depend on what's in the book... if the book is on programming, it isn't far removed from formal logic. For chip design, it would need to be rather encyclopedic, covering everything from EE basics, logic gate design, fundementals of chemistry, and the like... if it was done in a very blunt, step-by-step fashion, it could convey enough information to get something like a 4 bit counter/multiplier made from large galium arsenide crystals built using lampwork torches and blown glassware. From there its all practice and refinement though.
It would depend on the book.
Well, again, I think the only difference here is a difference of emphasis.
Books might help train the worker, but you need a worker with enough of that practice and enough of that refinement to actually make the minor iterative upgrades to the workshops that cumulatively enable better products over time.
You're just throwing the emphasis on the book, even while saying that it's practice and refinement that creates the actual change, while I'm saying it's practice and refinement that create the change, and books just help bridge the gap in practice sooner. That's still basically the same argument.
Metallurgy X could simply be a more advanced forge, and refinery, and then the rest would be little steps in between, or the work of 7 dwarves over a few years piling improvements on already established workshops. [the latter being a more progressive alternative to straight-on jumps with each tier of technology, again with their own limits, since grandpa's workshop can only be so fast, or hot without a complete overhaul]
The rest can be offloaded on the actual potentials/talents/affinities that your citizens can have, and society-based bonuses from having a pretty sizeable metalworking environment. [etc etc, it's expensive, and sieges swerve your curve, and you have to start over]
I agree with this as well - especially if we can get dwarves just doing passive upgrades on their workshops, we can have much more nuanced "tech levels", especially when it just becomes what can functionally also be passively modeling the concept of economy of scale.
The more this is discussed, the more I think that having truly difficult-to-build workshops as a requirement for building "higher-tech" items is a good idea.
I remember the "minimalist challenge" that involves embarking with nothing but a couple bars of copper and an anvil, where you can make one chunk of copper into a kiln, disassemble the wagon, burn that wood for charcoal, disassemble the kiln, make a forge, turn the copper into a pick, mine a hunk of copper or iron ore, and turn that into an axe for more wood for charcoal...
Once you have a single stone boulder, carved with nothing but a single crude copper pick, you can make what are the most technologically advanced workshops you will ever have access to. You can make minecarts from a log without any tools, for example.
The simple fact that workshops are so easy to build and disassemble makes them utterly transient, when we could be making them into something that have real weight and meaning in the game.
If those workshops can't do anything but the most basic, minimalistic jobs using nothing but a stone counter, (just enough to bootstrap your industry from Robinson Crusoe levels, but not enough for improved products,) and you have to pimp your workshop to get somewhere, then they become much more solid investments. (Of course, if you maintain your tech level, and you can still "take back" any item that went into a workshop without it being destroyed, you can still disassemble the workshops from one spot and build it back up in another, but it would at least be a much more laborious process, and a much greater sink of resources.)
Part of the question is, however, how do we do this level of increasingly refined tool-making?
Are we going to make it so that nothing more advanced than basic goods are available at the start, and to get -fine- goods, you need to upgrade in at least some way? Even then, is that enough?
Since there's nothing stopping an in-reality-pathetically-inadequate crude copper pick from mining through granite (which is hard enough that most real-world miners just reach for the TNT to get through) then does a limit on item quality have a real meaning for almost any product of worth besides for trading purposes?
Sure, we could block off access to more advanced versions of goods - glass doors or tables unavailable until you make a glassblower's forge with at least 10 upgrades, or soemthing, but since we're currently at a level in the game where anything past bare subsistence is just fluff...
The other alternative - that we have named iterations of items that have actual differences, is something that I'd honestly prefer to minimize to the greatest extent possible. Even with something like a *improved mica mechanism* and a -supreme steel mechanism- (provided having adjectives and quality indicators that convey entirely different things isn't confusing enough to start with...) rather than whole other names, we could still be looking at suddenly quintupling the number of items that we will be looking at in the stocks screen.