Newton's theory may be correct, but it just doesn't apply to firearms
The conservation of linear momentum applies
EXACTLY to firearms. Momentum - the product of mass and velocity - will be conserved. In fact, the momentum of a gun in recoil will be slightly
greater than the momentum imparted to a target because a) the bullet will slow down before hitting the target, due to aerodynamic drag forces, and b) the powder gasses from the propellent will impart momentum to the gun, but not the target. (Let's ignore muzzle brakes and muzzle-contact shots here.)
I remember testing this years ago - we took an 18" piece of railroad tie, stood it on end on a hard surface, and shot into it from about 3 feet away with a full power .41 Magnum. The wood weighed less than any man, but the bullet - which didn't penetrate all the way through - barely rocked it. Conventional "wisdom" would have had the wood blasted back ten feet or more, but it wasn't.
When a bullet hits a man or animal and a violent reaction is observed, it's because of the nervous system's reaction to the trauma. Think about it - if you're sitting on a barstool and someone sneaks up behind you and sticks you with a sewing needle, can you say the needle "knocked you down" if your legs get tangled in the stool's legs when you try to jump up and you end up on the floor?
For those of us who managed to complete compulsory education without taking a physics class, is there a measurement standard for momentum that can be applied to firearm projectiles? For example, if I know my 230 grain projectile is travelling at 800 fps when it strikes a stationary target, say a 25 lb. steel popper, can I predict how much momentum will be imparted to the target?
Assume it's an inelastic collision, i.e, the bullet sticks to the target and doesn't pass through. The momentum of the bullet before it hits will
exactly match the momentum of the system after the hit. Do a search on "ballistic pendulum" and you'll see lots of info on this. (Ballistic pendulums were used 'way back in blackpowder days to compute bullet velocities.)