I did a very quick check last night and I've taken the two greyscales attached. They were both taken within minutes of each other, nothing changed between runs apart from physically removing the lens. I left the aspect in my Edge set the same (4:3 for 16:9 squeeze) for this. I used my i1-LT with diffuser facing the screen having first created an offset using measurements off the screen. I'm using Chromapure Version 2.03 Professional. The projector is a JVC HD350, using a VideoEQ Pro for greyscale, gamma and gamut adjustments. The lens is an Isco II.
I think that's covered the technicallities. I know the i1-LT isn't the most accurate sensor (plus mine is about 2 years old), but it is the comparative results this is designed tos show rather than the absolute. I'm going to rent an i1 Pro in the near future, so if I remember I'll do this same test again then.
In brief the greyscale was slightly effected showing less red without the lens. The gamma was the same (within the bounds of repeatability) and there was a slight change in the measured colour gamut.
I hope this is of interest...I can't really tell any difference myself when watching with or without the lens in terms of colour difference/greyscale, so I don't bother doing a calibration for both cases. I do make any calibration adjustments with the lens in place however, as this is my typical viewing situation.
The left hand image is with lens and the right hand is without in each pair.