The search for a well-founded physical cosmology is ambitious, to say the least, and the heavy reliance on philosophy not encouraging. When Einstein (1917) adopted the assumption that the universe is close to homogeneous and isotropic, it was quite contrary to the available astronomical evidence; Klein (1966) was right to seek alternatives. But the observations now strongly support Einstein's homogeneity. Here is a case where philosophy led us to an aspect of physical reality, even though we don't know how to interpret the philosophy. Other aesthetic considerations have been less successful. The steady-state cosmology and the Einstein-de Sitter relativistic cosmology are elegant, but inconsistent with the observational evidence we have now.
The program of empirical tests of cosmology has been a productive research activity for seven decades. The results have greatly narrowed the options for a viable cosmology, and show that the relativistic Friedmann-Lemaître model passes impressively demanding checks. My survey of the state of the tests four years ago, in a Klein lecture, was organized around Table 1. I discuss here the provenance of this table, and how it would have to be revised to fit the present situation.
The table is crowded, in part because I tried to refer to the main open issues; there were lots of them. Some are resolved, but an updated table would be even more crowded, to reflect the considerable advances in developing new lines of evidence. This greatly enlarges the checks for consistency that are the key to establishing any element of physical science. Abundant checks are particularly important here, because astronomical evidence is limited, by what Nature chooses to show us and by our natural optimism in interpreting it.