The tests found that major differences, especially in dynamic modulus, exist in the different mixtures of asphalt, and even among samples of the same mixture. The results suggest that that the constitutive makeup of paving materials interact to have a greater effect on the performance of the material than might otherwise be suspected. In addition, it was found that certain results of the dynamic modulus tests were not accurately predicted by the current models, reflecting a need for greater accuracy in these models.

Personal Opinion:

The amount of difference in performance of paving material that can be created by relatively small differences in the makeup of that material was fairly surprising. Equally surprising -- and more dismaying -- is the fact that the current models missed...

The MEPDG was certainly a necessary implementation; one wonders what other states are as behind in their testing procedures. More detailed research -- such as the testing of a wide array of controlled mixtures and individual mixture constituents in their response to various stress tests -- would yield even clearer guidelines for testing and improved efficiency in the cost of paving and maintenance.
Though the paper puts itself forth as only an "initial effort" in the testing f hot-mix asphalt materials, I feel like perhaps more results could have been derived from the testing than were presented here. A few additional test would not have been too difficult, either, and the additional data provided might have allowed for broader implications to be derived from this study.