Recent working meeting with Richard Bower, Hannah Smithson, Tom McLeish and Brian Tanner worked through the surprisingly subtle physics issues of balancing luminous drag and absorption as the celestial spheres crystallise out.
Another surprise is the strong effect of the initial matter distribution (following the original expansion). Well behaved universes of the Aristotelian type seem to require steep initial density distributions, though this requirement may be balanced by sufficient absorption within the shells. Would Grosseteste be surprised to hear that for his universe to work, the heavenly spheres need to possess only partial transparency to “lumen”?
Somewhere in there is not just a medieval universe, but a medieval “multiverse”. All possible numbers of planetery spheres – and other universes where the density goes chaotic and non-monotonic. The initial density profile looks important too … more when the movie comes out!