I think you can do what you want by looking at each split leading to any leaf you desire and then setting up a data filter for your data table using those split criteria in the proper order as shown in the tree leading to the leaf. Select the data points that have been filtered and you have the points located in that particular leaf.
Thank you very much smoore2! Your answer makes sense. However, I am looking for a way to automate it. Since my tree will probably have 20 features 2.5 million rows of observations and fine tuning this tree will be very interactive process. So after each change of tree parameters, I will need to construct manual filters. It would be so much easier to see final leaf for each observation right away.
I kind of knew that would be your response! Perhaps in a future version of JMP, one could click on the leaf in the final tree and the filtering and/or selection could be automatic!?!?