Considering modern echinoderms together doesn’t work very well - you have to consider what you’d be likely to find in a single ancient echinoderm, which you are doing to some extent, but sort of ignoring fossils from the time. And projections with a superficial resemblance to pentaradial symmetry isn’t a great link on its own :
But note I’m not bringing up a comparison to velvet worms in general, but specifically to Orstenotubulus. That’s known in enough detail that a relationship to other lobopods is much more likely than echinoderms, and it has spines like Hallucigenia, making it a pretty good connection between those two. Your mileage may vary, but that seems much more right to me.
