"... the effect of dust on light propagation ... simply doesn't explain the observations."
But no term at all for this is included in the analysis. Long nano fibers would have an unfocused influence on the measurements.
"... why [do] you think the mathematical model is an "embarrassment" or are you simply trolling?"
The cosmological constant and many other representations of dark energy are ruled out by the Bianchi identities that apply. This becomes plain when when the proper tensor rank is used for the spacetime entities invoked. The cosmological constant is an artifact of using a second rank tensor to represent the Einstein tensor, and a simple divergence operation to show conservation, instead of using the fourth rank tensor and exterior derivative operation that satisfy the criterion of general covariance. The Bianchi identities translate to the conservation laws in the context of mechanics, but in geometry they are theorems which follow when the metric exists.
"... phenomenology ... "
General relativity is not a mere metaphor, as epicycles were.
" ... negative mass in Newton's formulae give antigravity ... "
Newton's formula for gravity does agree with this possibility that inheres in the Einstein tensor. It is also interesting that a simple acceleration of expansion gives supernovae that are closer and brighter at a given redshift.
"Yeah, yeah, tachyons with negative mass."
I should have written negative momentum, which is a better translation into mechanics of the possibilities allowed by the Einstein tensor. There is no principle in geometry which disallows this feature of the Einstein tensor. The Bianchi identities conserve superluminal momentum in that same superluminal direction. The identities even allow superluminal dilation horizons.
Spacial dilation causes divergent spacial curvature, which progressively weakens the intensity of light as it travels. This agrees with the phenomenon.
But a non conserved version of dark energy is contrary to the Bianchi identities.