The LGM ice edge in the Bristol Channel

Two of the maps from Giglio et al, 2022, representing recent thinking, in tune with the BRITICE-CHRONO work.  At the eastern edge of these maps there is a problem..........

 I have been reading an impressive article by Christiana Giglio et al (2022) about the evidence for ice advance and retreat during the late Devensian.  Details:

"Character of advance and retreat of the southwest sector of the British-Irish Ice Sheet during the last glaciation"
Cristiana Giglio et al 

Quaternary Science Reviews 291 (2022) 107655

It's an interesting and worthy contribution to our understanding of what went on in the Celtic Sea arena, with much information coming from studies of sea bed sediments and features off the SE coast of Ireland. The authors agree that the maximum extent of the LGM Irish Sea Ice Stream was right out on the Celtic Sea shelf edge, around 26,000 years ago.  The isochrones show the ice edge retreating at a phenomenal rate, clearing through the St George's Channel within 2,000 years of the maximum extent. Sea floor bedforms and "palaeo channels" seem to make sense almost everywhere, except for the eastern edge of the ice lobe.

Now (not for the first time) I have to have a little gripe, since the outermost eastern ice edge as shown does not make sense.  Nor is it supported by any evidence. For the umpteenth time, ice that is flowing in unconstrained situations does not flow parallel to the ice edge.  It flows perpendicular to it.  Why do people persistently fail to appreciate that point?  Part of the problem is the persistence in the literature of the South Pembrokeshire ice-free enclave or corridor, which we can see on both of the above maps.  Well, following the publication of my 2023 article in Quaternary Newsletter, we can now abandon the idea of an ice-free S Pembs during the LGM:

https://www.qra.org.uk/mp-files/qn158_1_late-devensian-ice-free-corridor-in-pembrokshire.pdf/

The ice on the western flank of the Irish Sea Ice Stream (ISIS) did flow parallel to the ice edge (or ice contact zone) because it was constrained and influenced by ice coming from the Irish Ice Cap.  That makes perfect sense.

By removing the considerable South Pembrokeshire constraint, which has caused researchers no end of problems over the years, we can accept that the LGM ice flow at the eastern edge of the ISIS was from the NW towards the SE, and not from the NE towards the SW.  This would be in tune with the alignment of the palaeo-channels mapped off the coast of Cornwall.  It would also fit glaciological theory!  It would make sense for the Isles of Scilly to have been a nunatak at a time of ice cover up to the Cornish coast, with the ice edge perhaps in places a little distance inland.  The other offshore islands (Ramsey, Skokholm, Skomer, Caldey and Lundy) were probably all ice-covered for a few centuries.