You're quite right -- the value 4 for the distance propagation factor is only an empirical approximation to reality. It varies from 2 for very short distances where propagation is line of sight up to more than 4 in some cases. Lee's other book (Mobile Communications Design Fundamentals, ISBN 0-672-22305-8) cites a variety of experiment-based models giving a range of distance exponents: 3.84 for a generic "surburban" area, 3.68 for Philadelpha PA and 4.31 for Newark NJ. Japanese studies show greater attenuation (particularly in Tokyo), probably because of that country's denser cities. The models are quite detailed and include antenna height and operating frequency as parameters.
I do know that at work everybody seems to use