A recent article in Nature reports on a large collaborative study done by 36 labs across the globe to examine the reproduciblity of the results of several classic and contemporary studies in psychology. It's reported that they could successfully reproduce the results of 10 out of 13 past experiments. Further details are published in an article on "Investigating variation in replicability: The “Many Labs” Replication Project".
The question remains open for reproduciblity of some of the observed travel behavior phenomena. It would be interesting to test, for example, the shape of the response function to waiting time, or existence of thresholds in price sensitivity, etc. (Idea: courtesy of Hani Mahmassani).
With regard to temporal or spatial transferability of travel behavior models, there are numerous studies in the literature. Perhaps the following dissertation by Sujan Sikder (University of South Florida) provides a comprehensive background on the transferability issue: Spatial Transferability of Activity-Based Travel