Clearance is %, but in the formula it is taken as a decimal. For example, 5% is taken as 0.05.
"n" values change with T,P. That's the reason I took 1.2 as an average for ballpark estimates. Values for ethylene are given in the technical literature.
For each stage they wouldn't change too much, and an average of "in and out" for each stage would suffice. There are other factors such as leakages, friction drops, changes in temperature, losses due to imperfect valve action, that decrease the real efficiency. By splitting the total compression into three stages you actually improve the VE, the first stage would have a higher VE than that of a single-stage. See the formula. I've seen a graph showing actual values of VE for air compression estimated for one-stage units as function of PR and % clearance that includes the other factors.
Since I cannot reproduce the graph, I'll give some values that will enable your own plotting of VE% as f(PR) the straight lines are the % clearances:
VE,% PR clea.,%
0.96 3 0
0.945 4 0
94 3 2
90.5 3 4
88 3 6
91 4 2
87 4 4
82.5 4 6
One notices that even for 0% clearance the VEs are not 100%.
n in a piston compressor is always lower than Cp/Cv. For example, for air the conventional Cp/Cv=1.4, however, actual polytropic values are 1.3-1.35.