I doubt that there is a simple answer to your question. Without going into pages, the problems with UV fluorescence as an analytic tool include the following:
- lamps vary somewhat in the wavelengths that they emit, and consequently the fluorescent effects that they have. Expensive ones for mineralogy costing hundreds of pounds and cheap ones for stamp collectors costing £15 or so do not necessarily act in the same way. A (very) rough analogy is the differences you get with incandescent bulbs, LEDS, or sunlight when taking pictures using visible light.
- human eyes vary, and colours are subjective. Some people cannot even distinguish red and green, remember. Is 'dusky pink' or 'pale straw' colour something we all recognise and agree on? I doubt it!
- the fluorescence from paperweights can vary depending upon impurities in the glass (and in turn the source of the raw materials), and even depend upon how the paperweights were annealed or held in the glory hole.
This is not to say that UV testing has no merits: for example, it is useful for showing whether there is lead in the glass, as with any UV lamp you will get a blue fluorescence with short wave radiation. And it is useful for helping attribute weights when there is a big enough database of results, and sufficiently great differences in fluorescence, to make tests meaningful (Kev H has shown this with Perthshire, Ysart Bros and so on). But even with many results there can be problems. I have been told by several people that antique Saint-Louis paperweights fluoresce 'dusky pink'. None of mine do under my UV lamp, at short or long wavelengths. That is not to say that the others are wrong - just that results can vary a lot, and that UV is by no means a magic bullet for solving problems of attribution.
What would help would be a comprehensive study using University laboratory standard equipment for illumination at specific UV wavelengths, and accurate spectrometry of the emitted light. But I doubt anyone has the money for that!