Hello
Tempering does *not* make steel harder and more brittle, it's the reverse! Here is how it works:
Step 1: Quenching (a form of harderning): The steel object is heated to a high level, then cooled very quickly by dipping in a liquid (water, oil, depends on what you are doing exactly). This has the effect of making the steel very hard, but also brittle.
Step 2: Tempering. After step one, the item is heated again, but *not that hot* - it could be done in a household stove. This slightly reverses the process, but overall it's a gain - you lose a bit of hardness, but lose a lot of the brittleness. A proper sword will be quench-hardened, then tempered.
Speaking of steel and sword, I'm not familiar with the properties of all the 400 series steel. However, I *can* tell you that "ordinary" stainless steel (420) is *not* good sword-making steel. It's too brittle! (it's great for knives though). A sword needs to be able to flex a bit. Spring steel is far better - in fact, swordsmiths of modern era with limited means have been know to make pretty good swords using pieces of a car's leaf spring.
I will note that I'm not a smith, but I was friends with one (unfortunately, he passed away recently) and I picked up a few things
Stainless steel is garbage for weapons and armor, not great for knives either because it doesn't hold an edge very well. It's used for cutlery because it doesn't do the nasty rusting thing and because obviously don't do much heavy duty work.
Dwarf technology level steel would be similar to late medieval blast furnace steel (the reaction implies a blast furnace, and Toady mentioned keeping the tech level <15th century), meaning a non-homogeneous, steel with various possible heat treatments, a somewhat streaked pattern under a microscope and various slag inclusions, mostly silicates, depending on the quality and origin.
The smith's skill was mostly keeping the steel as uniform as possible and getting the heat treatment just right. Making a piece of armor and
not heat treating it would actually be fairly daft and imply the skill doesn't know what he's doing, so I always assume that's where the quality levels come in. Various techniques were used, some used water-quenching and then tempered, others used oil quenching and more have been lost in the mist of time. All heat treatment is done after the forging though. It would be really silly to heat treat a bar of steel and then just heat it back up and forge it.
So yeah, trying to put numbers to end result of a medieval armorer is like trying to say how sweet apples are. Surviving specimens are also incredably rare (most are actually early renascence) and curators are not too giddy about them being put through stress tests. Though if someone knows of at least an approximation, if only from a reproduction piece, I would sure like to see it!
Though I wouldn't worry about it too much. To my knowledge adding a reaction to switch around an item's materials required DF hack, and as Urist Da Vinci said putting realistic numbers in DF doesn't guarantee realistic behaviours anyway.