Void demographics treat the empty regions as a population: the sizes, depths, shapes, and abundance of cosmic voids, compared against what ΛCDM simulations grow from Gaussian initial conditions (Ryden 1995; Platen et al. 2007). The model does produce a foam of voids and filaments qualitatively like the observed web, so the test lives in the statistics rather than the picture.
The statistics keep running empty-heavy. Surveys find more large, nearly empty voids than dark-matter-only simulations produce, with supervoids beyond 200 Mpc roughly five times more abundant than predicted, alongside depths and environmental dependences that require fine-tuned feedback or bias prescriptions to imitate (Tavasoli et al. 2013; Sutter et al. 2014). The structural problem mirrors the giant-wall problem in reverse: voids in the standard model grow by gradual gravitational evacuation, and the evacuation rate caps how large and how empty a void can become in the time available. A population of oversized, over-evacuated voids therefore points at the same place the oversized walls point, toward initial conditions or mechanisms with more large-scale organization than Gaussian noise grown by gravity.
The standing is a persistent population-level excess rather than a single disputed object, which makes it harder to dismiss and easier to measure. DESI, Euclid, and SKA void censuses will fix the void size function across enormous volumes, turning the factor-of-five claim into a precision statistic.