Ugh. Ballistics...
<mordin>
Couple notes here. The atmospheric drag is always highest just as the bullet exits the muzzle. From that moment on, the velocity of the projectile starts to slow down quite rapidly, and as atmospheric drag increases exponentially as velocity increases. That means the bullet slows down very rapidly after it's fired, and then settles on a downward arc falling at terminal velocity.
The projectile's impact angle will always be higher than firing angle. How much so, is determined by the projectile's mass, drag coefficient and muzzle velocity.
If you fire a BB gun at 45 degrees angle, I'd say the projectile will fall to the ground at angle between 90-75 degrees.
If you fire a cannon at 45 degrees angle, the projectile will likely impact ground at somewhere between 50-60 degrees.
As rifle rounds can have significant variation in their mass and aerodynamic properties, there can be significant differences in how exactly the bullet will fall down. A small caliber, light bullet with high muzzle velocity (such as 5.56x45 NATO) will start fast, go quite high quite quickly, but its velocity will slow down near the apex of its trajectory and then it'll start to fall down in an increasingly vertical path. As a lighter projectile, it will also have slower terminal velocity than a heavier bullet.
Something like 7.62x39 bullet will have slower muzzle velocity, so it'll not go as high, but it might actually go further because it doesn't slow down as fast at long range due to heavier projectile. It'll also fall faster. Something like 7.62x54R will fall faster still, and go longer with identical muzzle velocity.
So, as a summation one could say that the heavier the projectile, the better it will maintain its horizontal velocity component and the more dangerous it will be when it impacts ground.
The rifling provides another aspect to the problem. Rifled bullets have a spin (not the quantum sort) that stabilizes their direction by maintaining the bullet's attitude in flight. In other words, gyroscopic forces keep the bullet's tip pointed at one direction (excluding precession effects).
However, this is mainly useful when the bullet travels in a more or less straight line, to the direction it is fired at. Rifling therefore is a great advantage with small arms that are usually fired nearly horizontally. However, as the bullet is fired at upward angle and it embarks upon a ballistic trajectory, the spin from the rifling will try to maintain its velocity even as it reaches the top of the flight path. Therefore, the bullet will start falling sideways while still spinning. This'll cause an aerodynamic force perpendicular to the bullet's path (this is the same thing that enables curve balls in football, baseball and other sports; it's called Magnus effect and is caused by an effect called circulation and technically it has the same effect as an aerofoil by generating lift, search it on wiki if you want more information) so the bullet's path starts to curve sideways while it falls down and the rotation slows down.
This can also cause very interesting displays with ricocheting tracer rounds - as they tumble through the air, they can fly in a spiral path, spin wildly, or sometimes just fly nicely in a parabolic arc.
Thus, technically, a round fired straight up will maintain its aerodynamic stability better than one fired at high to medium angle. If it doesn't stay in the air long enough that the spin dies down completely, the bullet will start falling with its bottom first, and this position is much stabler than a non-spinning, tumbling bullet, so it'll have faster terminal velocity.
Of course, an aerodynamically stable projectile would, after its spin slowed down, stay pointed at direction of relative wind by aerodynamic forces, and a bullet like this would be the most dangerous while falling. I believe certain boat-tail bullets - might fall relatively straight, but who knows.
To ascertain the actual terminal velocities of different rifle caliber projectiles fired from different weapons, empirical testing would be required.
As to what types of injuries they could cause - I'm reasonably sure that small arms projectiles could cause bruising and lacerations, but most likely they would not have enough impact energy to cause dangerous perforation injuries or fractures.
If you were unlucky enough to get one fall in your face, sure you could get a wound or punctured eye. If one falls on top of your dome, I don't think it would have enough energy to cause a skull fracture - it would sting and you would probably get a bruise, but I don't think it would be likely to cause significant head injury.
This is based on speculation and gut feeling though, so don't take my word for it. Again, empirical testing would be required to gain proper results. Ballistic jello covered terrain would be the solution to this, while you fire different projectiles on it and hope they fall on the jello.
However, something like falling down AAA-munitions (like 12.7mm or 20mm AP rounds, never mind 30mm HE) are quite a different thing and I would want to be as far from them as possible.
</mordin>
