Interesting. The lack of large fossil remains suggests that not only was the head shark-like, but the skeleton was, too, in that it must have been cartillaginous, not bony (otherwise, you might expect to find more substantial fossils). Also, head evolution was always going to be more reliable than scales and spines as a guide to evolution, because of what heads tell about the brain, and because the brain is less wont to sudden and dramatic evolutionary changes than scales or spines. Ironically, by similar reasoning, Acanthodes are more likely to have evolved from a lamprey-like agnathan than from a bony-headed agnathan (the ostracoderms), because lampreys also have a cartillaginous skeleton, and only have a little bit of bone-like material (the rasping tongue).