@oralloy,
1 Your use of (mostly) EU time sequence names( like Cisuralian and Guadelupian) made me smile. Ive been talking North Amreicn age nomenclature and the CGS/ USGS and GSA is about North America.(The ICS system is trying to base itself on Global Boundary Stratotype Sections and Points GSSP), As far as I know, its still a work-in -progress- experiment at making the next steps into "Bar coding" most geo names. So far they have trouble on where to hold meetings.
2. Your defaulting to "I understand that such begat such and such" Is kinda lumping everything together and there are many many new specimens of radiated species of pelycosaurs that provide us with the better roadmap to mammals.Each year we get hundreds of new specie fossils of the Permian alone
While pelycosaurs had some similarites of mammals (like differentiated teeth). there were a whole lot of things they failed to have as we make new finds and make us deal with the most confusing issues of derived structure, that is the various skull openings and from where they derived .
Pelycosaurs were reptiles not protomammals because there were lots of more derived species that rose from taphinocephalids and theriodonts and a new one whose name I cant remember now. pelycosaurs remained in the fossil record until the end of the early TRIASSIC, (mammals were already evolved .
The issue of common ancestry precluded by contemporeneity is pretty much established. We keep opening the evidence pile with more species and it makes the confusion even deeper .(Most of the findings come from hobbyist fossil hunters whose findings may not be classified for decades )
I use these critters for locating strat units in very small time sequences because associated deposits of some specific things I look for occur as a function of a time sequence in which dimetrodons occur. Theyre a tool. When I teach applied strat, I have paleo students as well as applied science geos who will be working for mining rights and we carefully need to separate a dimetrodon from a moschops or a finless thrinaxodon or an evolving kannemeyeria.
Ill beg off the issue based on Steve Stanley's "distribution of lifeforms" My use of "lizard" for dimetrodon is based on the fact of its group name pelycosaur (Bowl lizard) The fact that its group remained till the Triassic also means that youve gotta look more carefully for the real common ancestor of Morganucodon or other "lizard Rats".
The arguments about "what do we call this thing" are kind of old fashioned, biostratigraphy and evolutio n are more interested in derivation and "Finely linking" species.