Steve,
From the report:
For the purpose of the analysis, the coefficient of friction can be generalized to encompass, without distinction, the static, dynamic, and rolling coefficients of friction that prevail during the putt of a golf ball starting at rest and rolling to a stop. It can be normalized to an all-inclusive parameter because of its dependence on many variables. Among them the most influential of which would be the height of cut, are the morphological and growing characteristics of the turfgrass species, the turf density and uniformity, the thatch layer, the dimpling pattern and the construction of the golf ball, the season, the wetness, even the time of day. Despite the influence of these variables and others, the green speeds of "as built" undulating greens can, with reasonable accuracy, be articulated and prepared analogous to the benchmark green speed indices from Stimpmeter measurements taken on level greens.
So it does seem that a reasonably good number could be derived that would answer David's question.