I have used this exercise several times, but I am reporting the results from the Fall 2017 semester.
Students accessed the structures, measured the bond angles using Mercury, and calculated the tau4' values without any difficulties (questions 1 and 2).
When they got to the third question, they could describe what they observed, but struggled with the language. They were very concerned about how to name the observed structures. They were not satisfied with using the terms "distorted square planar" and "distorted tetrahedral" to describe the structures. (This then led into the discussion of the tau4' values and why focusing on the names of the strucutres was limiting.)
All of my students were also able to calculate the LFSE values for the Ni(II) center in the four geometries. They asked about the spin state, but I prodded them to talk it through themselves and think back to previous discussions. They quickly realized that for some of the geometries there is no difference between the HS and LS configurations. They decided to calculate the LFSE for both configuations when they were different. Once their calculations were complete, the students determined that square planar should be the preferred geometry based upon the LFSE.
The last question is the one that threw a monkey wrench into what they thought they knew. They were surprised that a d8 metal center would adopt a tetrahedral geometry since this was contrary to what they had originally learned. I then asked about what other influences would impact the observed geometry. About half of my students said that the steric repulsion of the four donor atoms (and other atoms in the tropocoronand ligand) in a square planar arrangement was greater than that in a tetrahedral arrangement. These students were then able to make the connection to the fact that this must outweigh the LFSE value and favor the geometric transition of the nickel center.