Firstly, you have a single pole roll off as the graphs you have supplied give a 45 degree phase difference at the -3dB point. The roll off is a function of the amplifier's internal gain, so although you will get a phase shift in the output that results in a phase shift in the feedback loop, it will not get amplified because the amplifier does not have enough gain to amplify it (it is losing gain due to the internal gain roll off). If it increases slightly in the lab, you probably have some stray capacitance (scope probe?) that is causing an additional phase shift and distorting the natural single pole roll off that you are seeing in your LT SPICE simulation.
Feedback analysis of op amps is tricky, but I think this is what you are seeing