Dear cos-top friends,
It seems that not only is there at least some interaction between cosmic topology and gravity - the residual gravity acceleration effect as derived heuristically for the 1-torus and 3-torus[1], but the effect distinguishes 3 well-proportioned spaces (T^3, S^3/T^*, S^3/O^*) as special in that the effect cancels to third order in the Taylor expansion, and even more interesting, the Poincare dodecahedral space cancels even better than these other three well-proportioned spaces. The effect in S^3/I^* cancels down to fifth order in the Taylor expansion![2] In this sense, it's better balanced than the other spaces. Maybe we could say that the former three spaces are well-proportioned and "well-balanced", while the Poincare space is well-proportioned and "super-well-balanced"?
This dynamical/geometrical/topological result is independent of the empirical arguments in favour of the Poincare space.
Enjoy! :) boud
PS: There'll be a cosmic topology session at the 12-th Marcel Grossmann Meeting in Paris, 12-18 July 2009. http://www.icra.it/MG/mg12/en/ Contact Marek Demianski and register if you're interested.
[1] RBBSJ, 2007, http://arxiv.org/abs/astro-ph/0602159 Astron.Astrophys.463:861 [2] RR 2009, http://arxiv.org/abs/0902.3402, subm A&A