Try setting the atom size to 2, the number of atoms to 1000, and the gravitational constant to 0.02. Then the lapse-rate delta-T over the height of the container is 2.5, which would show up easily. Adjust the system's total energy (using the Faster and Slower buttons) until you see a good strong density gradient; an average temperature of 1.5 or so works well. After the system has equilibrated, click the "Save state" button to get a snapshot of all the particles' positions and velocities. Copy this into a spreadsheet and analyze. I binned the particles by height into ten groups, plotted the average kinetic energy as a function of height, and found no statistically significant dependence at all. Again, the lapse rate delta-T would be 2.5 which would have been obvious.
Now I'm wondering how, exactly, you COULD simulate a system that exhibits the adiabatic lapse rate.