Hmm, I think my report was a MS Works file.... so that's lost forever. To bad I hadn't discovered LaTeX by then.
Did find some random data on the subject:
==> Spread
Measured in units of D, where D is the diameter of the cross-hair
Spread
Throwing Knife 1
PP7 1.5
PP7 (Silenced) 1.5
DD44 Dostovei 2
KLOBB 6
KF7 Soviet 5
ZMG (9mm) 4
D5K Deutsche 3.25
US AR33 Assault 3
RC-P90 3
Automatic Shotgun 7.5
Sniper Rifle 1.5
Cougar Magnum 1.25
Golden Gun 2
MoonRaker Laser 1
Remote Mines NA
I think the report said the method used to collect that. Recalling from memory, it was something like standing is a set spot every time and firing at a 90deg plane to a fixed point on a nearby wall. To reset for the next round I'd just shoot elsewhere until the bullet holes would disappear.
Edit: Some on the auto-aim.... now again without the report this is based off memory, but it tested several scenarios. Std. Auto aim, Blind opponent (has no gun or your aiming across and invisible wall both of which disables auto-aim), Std. Auto-Aim while the shooter is crouched, Blind opponent while the shooter is crouched. Both subjects were placed as consistently as possible (wedged against walls for example) and auto-aim or lack there of did the rest.
Auto Aim / Blind / Auto Aim (Crouched) / Blind (Crouched)
Head Shot= 3 Head Shot= 12 Head Shot= 1 Head Shot= 0
Body Shot= 23 Body Shot= 17 Body Shot= 32 Body Shot= 0
Limb Shot= 24 Limb Shot= 21 Limb Shot= 17 Limb Shot= 50
So you'll actually get more random head shots without auto aim if you just stand (and just be shooting at legs if you crouch)