The original Palaeozoic global full-plate model (410-250 Ma) is that of Domeier & Torsvik (2014). Assuming that the TUZO and JASON LLSVPs have remained nearly stationary in pre-Pangea times, continents were reconstructed in latitude from palaeomagnetic data (e.g., 410±10 Ma Laurussia pole fitted to the geographic South Pole), and calibrated in longitude such that Large Igneous Provinces (LIPs) and kimberlites were positioned above the edges of TUZO and JASON at eruption time. This procedure is termed the plume generation zone (PGZ) reconstruction method (Torsvik et al. 2014) and the palaeomagnetic frame were corrected for True Polar Wander (TPW) using an iterative approach. In the diagram above, we show the starting point for our Late Palaeozoic model at 410 Ma: This is a palaeomagnetic full-plate reconstruction at 410 Ma where kimberlites and one LIP were used to calibrate continental longitudes for Laurussia (including Laurentia; Avalonia, red shading; Baltica, blue shading) and Siberia (yellow shading) at this specific time. The PGZs (red lines) of JASON and TUZO have been counter-rotated to account for TPW. Because we have estimated TPW in our model, we can theoretically use the TPW-corrected frame for forward mantle modelling, but note that the oceans (about 70% of the Earth’s surface) are essentially synthetic (“Made-up”).

References 
Domeier & Torsvik (2014): Plate tectonics in the late Paleozoic. Geoscience Frontiers 5, 303-350.
Matthews, K., Maloney, K.T., Zahirovic, S.,Williams, S.E., Seton, M. & Müller, R.D. (2016). Global plate boundary evolution and kinematics since the late Paleozoic. Global and Planetary Change 146, 226–250. 
Torsvik, T.H. & Van der Voo, R., 2002. Refining Gondwana and Pangea Palaeogeography: Estimates of Phanerozoic (octupole) non-dipole fields. Geophys. J. Intern., 151, 771-794. 
Torsvik, T. H. et al. (2012): Phanerozoic polar wander, paleogeography and dynamics. Earth Sci. Rev. 114, 325–368. 
Torsvik et al. (2014): Deep mantle structure as a reference frame for movements in and on the Earth. PNAS 111, 24, 8735–8740. 
Young, A., Flament, N., Maloney, K., Williams, S., Matthews, K., Zahirovic, S. & Müller, R.D. (2019). Global kinematics of tectonic plates and subduction zones since the late Paleozoic Era. Geoscience Frontiers 10, 989-1013. 

A global full-plate model by Matthews et al. (2016) for the past 410 Myrs is essentially a copy of our 2014 model between 410 and 250 Ma, and thus acknowledging our palaeomagnetic data selection and the PGZ method (assuming stability of JASON and TUZO) to calibrate continental longitudes. A more recent model, however, by Young et al. (2019), co-authored by many of the same authors as in the Matthews et al. (2016) paper, contains some severe flaws and should not be used for plate tectonic reconstructions nor forward mantle modelling.

Our original plate model is largely based on apparent polar wander paths (APWPs) for Gondwana, Siberia, Laurentia/Baltica (Laurussia after 430 Ma), and their later combinations into Pangea (Torsvik et al. 2012). At 410 Ma palaeomagnetic data from Laurussia is extremely well defined, but Young et al. (2018) moved Laurussia arbitrarily 30° northward. In addition, Siberia was also moved northwards because they kept the same relative fit between Laurussia and Siberia as in Domeier & Torsvik (2014). Both Laurussia (see mismatch of 410 Ma pole with the geographic south poles) and Siberia are therefore at clear odds with the palaeomagnetic data, i.e. the only data that can reconstruct continents in latitude before the Jurassic. They also moved Laurussia-Siberia arbitrarily 49° eastward, and Gondwana (based on an older APWP of Torsvik & Van der Voo, 2002) 11° westward compared with Domeier & Torsvik (2014).

Having used an old APWP for Gondwana (not corrected for potential inclination shallowing in clastic sediments, not corrected for TPW, and with arbitrary longitudes) together with flawed latitudinal reconstructions of Laurussia/Siberia (and arbitrary longitudes), Young et al. (2019) concluded from numerical mantle modelling exercises, starting at 410 Ma, that TUZO and JASON are unstable, and therefore the PGZ reconstruction method used in Domeier & Torsvik (2014) is invalid. Because many of the relative fits that Young et al. (2019) have preserved (for example the aforementioned Laurussia-Siberia fit) from our earlier model are in fact an outcome of our use of the PGZ method, their conclusions also invalidate their model, analysis, and indeed the conclusions themselves.