Many scenes do not change much over the NIR band. Thus it's not really surprising to observe that the main influence is getting more even distribution over the channels (meaning sooner or alter there is no red to swap for blue as both are virtually identical) because the Bayer dyes get increasingly trasparent. The sky will darken as well, if you are away from urban environments with their ever-prevailing atmospheric dust and haze.
I was surprised by the colour hues present in the UG1 capture, though. Not using this filter myself, only the B+W 403, that I thought would be more or less the same. Obviously it isn't.
When you do narrow-band IR stacks, sometimes you come across subjects for which the spectral proiperties change rapidly across NIR. Then you can get quite unexpected colour combinations. Such irregularities are rarely captured or recognised when one uses the normal broad long-pass IR filters.
