The 'questionable garage test' really does not do either the bows, crossbows or the armor any justice. A thin, flat sheet of mild steel strapped to a hay bale is not an analog for a knight, its an analog for a hay bale wearing a thin sheet of mild steel. They are the historian's equivalent of the 'Perfectly spherical horse in a vacuum'.
So, as my part in correcting common misconceptions and myths:
Plate armor was pretty much impervious to both crossbows and bows when shot at frontally. Actual historic accounts (ie Agincourt) make this pretty obvious, as do the more legitimate practical tests.
Mail still had a pretty impressive resistance to arrows and bolts, as can be figured from accounts of crusaders being turned into walking pincushions by their Saracen enemies. And contrary to popular belief, the Turks had fairly weighty bows of their own, of similar construction of other nomadic, central Asian tribes such as the Huns, Cumans and Mongols.
Nobody would have bothered with armor if it was as useless as modern media makes it look.
The same goes for the endless crossbows vs bows thing. They were both used extensively. If either of them was crap, they would not have bothered them.
Bows generally hit harder at long range as the arrows have a lower front profile/mass ratio, but long range shooting against armored targets was not really done in the first place as it was a waste of ammunition. If possible, archers went for the poor sods with poor armor, if not they waited for the enemy to get closer and shot at them with flat trajectories, not the piss-arch you often see in movies. That still didn't kill them, but still hurts and demoralizes them.
One thing about crossbows I haven't seen anyone bring up: They are quite a bit more convenient to use behind cover. A crossbowman can reload behind a crenelation, wall or pavise and then only needs to stick out his head and a bit of his shoulder to shoot. This is a very big advantage in long-term shootouts like sieges, which also happened to be significantly more common in the middle ages than field battles. Rate of fire becomes a lot less important when your battle can last more than a year, and your supply of ammunition is finite.
...
To get back to something more relevant to the actual game, raw modding hasn't gotten me all that far. All the things like damage fall-off, accuracy, rate of fire, skill scaling and AI behavior are completely inaccessible with the raws. Getting the game to differentiate between hunting and war bows/crossbows has also not gone well for me.
I'm hoping DFhack and future version of DF can get a bit further with this.