Tamlin,
I very strongly recommend that you get hold of a copy of the paper. It is really a review and discussion of the concept of fuzzy morphology, in which the boundaries between organs are blurred (as opposed to classical morphology, with discrete categories of roots, leaves, stems etc.), but it takes Utricularia as a case study in a lot of depth, with many illustrations of meristems etc.
To say that Utricularia lacks roots & leaves, and is therefore composed of modified stems/stolons is an oversimplification. They point out the similarity between Pinguicula roots (atypically lacking root caps) and stolons of eg. U. longifolia, which have a similar arrangement of vascular tissue. I don't know much about Genlisea, not growing it myself, but it could well be that the trap 'leaves' really have at least a partial origin in roots, and likewise for Utricularia 'stolons'. I accept that Utricularia does not have 'classical' roots & leaves, but it doesn't have 'classical' stolons either. My main point is that Utricularia have not 'lost' leaves & roots, they have rather merged their vegetative organs together to produce a morphology that doesn't fit into classical morphological boxes. This is found in many other species as well,to some extent, as in compound leaves which may have stem characteristics. Another example is the climbing fern Lygodium, where the 'leaves' have indeterminate growth like a stem, so it is impossible to distinguish separate leaves & stems.
Giles