curvature corrections less than 10% for r < 80 Mpc
Eq. (15) of 0710.5073 - 10% level at about 50Mpc, so above this the effects are relatively small
Fig. 2: this inhomogeneous modelling correctly fits the growth of the variance with z in the Freedman et al 2001 H_0 set of estimates, while variance calculated for a homogeneous model underestimates the variance
is the "peculiar gravitational potential", i.e. initial power spectrum, it does not have the radiation-matter turn over
idea: this method should be used when modelling H_0 in general, for the regions below about 150 Mpc
idea: some work done in this direction - consider how H_0 varies in different directions and model this inhomogeneously