It would be interesting to see that paper. One reason that I can think of, that doesn't involve mass transfer issues and cell geometries, is that in a 2-electrode cell, your counter is your reference, and its potential will change depending on the current (standard potential plus the overpotential necessary to run the required counter reaction). If you try to increase the current by increasing the voltage of the cell, that voltage increase will contribute not only to the overpotential on the working, as it would in a 3-electrode cell, but also to the overpotential on the counter. So your working will effectively "see" only part of the driving force and the current will be lower.

In a 3-electrode cell:

E_applied=(E_std_W + n_W)-(E_std_R)

In a 2-electrode cell:

E_applied=(E_std_W + n_W)-(E_std_C + n_C)

This is just a guess as to why you see what you see. I'd be curious to find out what you figure out.

In a typical 3 electrode setup the working potential is the voltage between the working electrode and the reference electrode. The counter electrode is set to whatever potential is required to pass the same current that the working electrode does. This means that the 3-electrode voltage and 2-electrode voltage are not directly comparable. However by measuring the voltage between the counter electrode and the reference electrode and the voltage between the working electrode and the reference electrode, one can use simple algebra to calculate the potential between the working and counter electrodes.

One must also be careful about the identity of the reference electrode--it is best to only compare measurements against the same reference electrode, but it is also possible to convert from one scale to another.