Truncating or windowing an IR cannot cause phase issues, and most IRs certainly do have truncated (or windowed if done correctly) early reflections. You will actually get less phase cancellations in a shorter IR because fewer reflections will be present. I think that in this case, calling them phase 'issues' may be confusing to people, because they will assume that if they hear an unpleasant artefact - they will assume that it is a phase cancellation, while in reality plenty of the sounds that we like contain lots of phase cancellations.
Closed back guitar cabs, or cabs with multiple drivers, for example, have lots of phase cancellations that make them sound how they do. As it has been pointed out, most guitar recordings have been recorded NF with a reflection from the floor that causes phase cancellations. This alters frequency response, and phase is the reason why. It doesn't mean it will sound bad. I cannot say what a reflection-free NF IR would sound like, because AFAIK none exist. It would be interesting to hear one. Even a test of an NF IR at various heights in order to hear the effect of attenuating the effects of the phase cancellation from the floor, at varied amounts, in an otherwise reflection free environment (to minimise other variables). I have the ability to do this test, but am not interested enough in NF IRs so I see it as a waste of valuable time (which goes by very fast during a FF IR capture session).