Yet another burst of data from my current favourite Type 1a supernovae has brought the spotlight back onto dark energy, reexposing one of those hidden assumptions that seem to have permeated cosmological thinking sans scrutiny.

It has long been written into the equations that gravitational attraction between galaxies should be slowing the Hubble expansion, as though observable space was embedded in some higher dimensional space through which Newton's inverse square law strangely persists.

Back in more local space, it is well known and mathematically provable, that a massive hollow sphere exhibits zero net attraction on any mass inside. At cosmological scales, the universe exhibits angular isotropy which must make the sum of distant galaxies gravitationally equivalent to a hollow sphere to the limits of accuracy.

So while gravity and dark matter no doubt play a vital role in cluster and supercluster formation, the idea that gravity is acting to slow the expansion might just be a nonsense that needs to be subtracted out of our dark energy models.