I probably will give it a shot, as time permits and need arises. As far as i see it, a conversion in general seems possible. The main problems (as far as i discovered them), in my opinion are:

Poser's 3-independent-axis posing system with euler angles vs. blender's quaternion posing. I.e. in Poser you can have different falloff zones for each of the x/y/z axes of a joint, in blender you typically have only one. Does the resulting figure have only one weightmap for the whole joint, or would you simulate the 3-axis bending? What does Interposer do in that case, btw? Also there is a question of whether to reorient rotation axes (blender uses a fixed "twist" axis) or to keep the order from poser (better recognition value if you know the poser figures), or use both.

Materials are very similar between Poser and blender at the time, so a conversion is possible, but if you want to render in cycles anyway (at least in some distant future you probably will), it is probably not worth the effort to convert the materials from one to another. At the current time, alas, cycles is not able to render some of the nice specials of blender, e.g. hair, so cycle's use is limited, so you might want to convert the materials for now to the blender materials, and then in two years throw it away to convert to cycles matials? Hm, much work...

Despite of the fact that each morph is rather easy to import, there is a display problem; In blender you have a single list of shapekeys, in Poser you typically have your morphs in a hierarchy that reflects the parent-child relationship of the parts of the figure, so you can select only the ones you need. With e.g. the Victoria++ morphs (very common), you get a huge list of several hundred morphs, and since blender uses a simple flat list for all shapekeys, you would get a large list of probably over 1000 shapekeys, which would be very uncomfortable to use. To make these usable, there is probably some kind of GUI creation involved.