This seems a simple enough question, but a moderately intense Google failed to find any simple answers. Since you're happy with a "rough estimate based on logic" I'd point out that car tyres are flat along a line normal to the direction of motion, so the displacement of water is going to be comparable to pressing a rectangular slab into a film of water.
The initial motion of the water is horizontal, but as the displaced water hits the stationary water outside the tyre contact patch the water will rise up in the same sort of way a wave breaks (a wave breaks because the water at the top is moving faster than the water at the bottom).
The velocity of the displaced water is simply related to the volume displaced per unit time, and therefore to the car velocity. The difficult bit is working out the mechanics of the "wave breaking" and how far the spray will go. A quick look at the Wikipedia article on wave breaking suggests this is far from a fully understood problem.
At a guess, I would say that above some critical speed most of the displaced water is simply displaced upwards with a small percentage being lost to viscous losses in the puddle. Ignoring air resistance the distance travelled by the spray will be roughly proportional to the car velocity, however I'd guess air resistance is important for water spray so the range will be proportional to velocity at low velocities but won't increase as fast for high velocities. I would also guess that the puddle depth affects the angle of travel of the spray, though I hesitate to guess the form of this relationship.