Li' correlation should use EoS critical volumes?

[longemen3000]

on Cubic EoS (and in general EoS, except models that fix the the complete critical point), the critical volume estimated by the EoS is different than the experimental critical volume, different that what is used here: https://github.com/moyner/MultiComponentFlash.jl/blob/0056d89e4c53b450448f01835a6f774d61c0a01a/src/utils.jl#L13 . i suppose that the differences aren't that big if you are far from the critical point, and this seems to be a correlation anyway, but maybe it could affect the labeling if you approach critical conditions?