According to the Wkipedia entry, there are a number of methods for calculating the COR.
The coefficient can also be found with:
COR = square root of h/H
for an object bouncing off a stationary object, such as a floor, where
h is the bounce height
H is the drop height
So I think this is the relevant measure. The fact that the ground deforms during impact is irrelevant. Here are some more details from the Wikipedia entry:
The COR is generally a number in the range [0,1]. Qualitatively, 1 represents a perfectly elastic collision, while 0 represents a perfectly inelastic collision. A COR greater than one is theoretically possible, representing a collision that generates kinetic energy, such as land mines being thrown together and exploding. For other examples, some recent studies have clarified that COR can take a value greater than one in a special case of oblique collisions[1][2][3]. These phenomena are due to the change of rebound trajectory of a ball caused by a soft target wall. A COR less than zero is also theoretically possible, representing a collision that pulls two objects closer together instead of bouncing them apart.
An important point: the COR is a property of a collision, not necessarily an object. For example, if you had 5 different types of objects colliding, you would have {5 \choose 2} = 10 different CORs (ignoring the possible ways and orientations in which the objects collide), one for each possible collision between any two object types.
Generally, the COR (e) is thought to be independent of collision speed. However, in a series of experiments performed at Florida State University in 1955, it was shown that the value of e varies as the collision speed approaches zero, first rising significantly as the speed drops, then dropping significantly as the speed drops to about 1 cm/sec and finally rising again as the collision speed approaches zero. This effect was observed in slow speed collisions involving a number of different metals.