Speaking of axes, I saw your testing, and found the differences very interesting. In particular that axes did so poorly against a giant elephant, while they excelled in the tests I ran.
I suspect that it's not just the strength, but also the material of the axe when trying to effectively chop into a beast as large and tough as a giant elephant.
Yeah, a steel axe wielded by a very strong dwarf probably penetrates at least twice as deeply.
When I did my giant elephant testing we didn't have DFHack yet. Now with DFHack everything including quality can be tested.
I just tried a simple test with a Goblin with Grand Master combat stats with elevated-physical , vs an Animated Dead Giant Elephant with elevated-physical 100 (so it's very bad in combat and is just a big dumb punching bag.)
I used 4 variants of weapon: noquality iron axe, masterwork iron axe, noquality steel axe, masterwork steel axe.
Unsurprisingly the masterwork steel axe gets the fastest kill, but this is followed by the masterwork iron axe which takes about thrice as long to get the kill. And then the noquality axes appear to be simply incapable of killing the zombie giant elephant, eventually of course the goblin would pass out from exhaustion then get killed, but this took about 5x longer than it took the masterwork iron axe to get the kill. I only did a few trials, but this was pretty consistent, and as I understand this kind of combat it's not that random, there are no quick decisive lucky victories, the goblin just has to hit the Giant Elephant in the head enough times that the skull runs out of "hitpoints" and can get cloven, and the number of hacks was an average of 7 with the masterwork steel axe, 34 with the masterwork iron axe, and the other two axes got about 170 hacks to the head before the goblin got exhausted, so it wasn't like the other two axes were just unluckily not hitting the zombie in the head.
So something about the quality is enabling the masterwork iron axe to destroy the zombie whereas the noquality steel axe isn't just worse, it can't do it at all! My preliminary speculation would be that it has to do with whatever benefit blunt weapons get from being higher quality, and the edged->blunt conversion, because I'm pretty sure it can't just be about sharpness, since my understanding is a no quality steel axe should have about the same penetration as a masterwork iron axe. It's also obviously not about the to-hit bonus, at least not if the to-hit bonus is only about overcoming defensive skills, since the goblins are hitting every time regardless.
edit: Having run analysis on 8 combat reloads, here's the number of head hacks required for the goblin to defeat, or die to, the zombie giant elephant:
| | Label | Mean ± SD | Head hacks |
|---------|---------------|-------------|------------------------------------------|
| Fight 1 | noq-iron-axe | 165 ± 44.1 | [168, 193, 167, 58, 189, 181, 183, 180] |
| Fight 2 | mas-iron-axe | 34.1 ± 10 | [19, 45, 32, 26, 36, 46, 26, 43] |
| Fight 3 | noq-steel-axe | 172 ± 18.2 | [165, 161, 196, 149, 188, 180, 186, 149] |
| Fight 4 | mas-steel-axe | 6.62 ± 3.16 | [5, 13, 4, 8, 4, 9, 5, 5] |
So I only did 8 trials, but it can be seen that there's a dramatic difference in the number of head hacks required for a kill.
While this an extreme test, I think we can conclude that quality is exceedingly important to account for.
edit2: Since having combat logs to analyze is amazing, here is how I do it. The DFHack script "enable devel/annc-monitor" will dump reports to the DFhack terminal, but it's more useful putting them in a file. For this, the script/devel/annc-monitor.lua has to be modified, modify the contents of "function log(s, color)" inserting or replacing with the following lines:
fd = io.open('reports.txt', 'a')
fd:write(s, '\n')
fd:close()
Thus writing announcements to a file "reports.txt" in the Dwarf Fortress folder.